oxygen sensitivity of skin neuroepithelial cells in
Post on 01-Mar-2022
2 Views
Preview:
TRANSCRIPT
OXYGEN SENSITIVITY OF SKIN
NEUROEPITHELIAL CELLS IN DEVELOPING
ZEBRAFISH, DANIO RERIO
by
Maria Louise Coccimiglio
Thesis submitted to the
Faculty of Graduate and Postdoctoral Studies
University of Ottawa
in partial fulfillment of the requirements for the
MSc Degree in the
Ottawa-Carleton Institute of Biology
November 15, 2011
© Maria Louise Coccimiglio, Ottawa, Canada, 2011
ii
ABSTRACT
In zebrafish, the ventilatory response to hypoxia first develops at 3 days post-fertilization
(d.p.f.) before O2-chemoreceptive neuroepithelial cells (NECs) of the gill appear at 7 d.p.f. This
indicates the presence of extrabranchial chemoreceptors in embryos and a developmental
transition to primarily gill O2 sensing. This thesis examined the skin NECs, which reach peak
density in embryos but decline as gill NECs appear. Exposure of embryos and larvae to chronic
hypoxia prevented the loss of skin NECs, shifted peak basal ventilation to a later developmental
stage, and induced a hypoventilatory response to acute hypoxia. Chronic exposure to hyperoxia
rapidly diminished skin NECs, shifted peak ventilation to earlier stages and eliminated the
response to acute hypoxia. Administration of the neurotoxin 6-hydroxydopamine degraded
nerve terminals that contact skin NECs and reduced both basal ventilation frequency and the
hypoxic ventilatory response. Thus, skin NECs are candidates for extrabranchial O2
chemoreceptors in developing zebrafish.
iii
RÉSUMÉ
Chez le poisson-zèbre, la réponse ventilatoire à l'hypoxie se développe à partir de 3 jours
post-fécondation (d.p.f.), avant l’apparition, à 7 dpf, des cellules neuro-épithéliales (CNE)
chémoreceptives d'O2 des branchies. Ceci suggère la présence de chémorécepteurs extra-
branchiaux chez les embryons et une transition développementale vers une détection de l’O2
principalement par les branchies. Cette thèse porte sur les CNE de la peau, qui atteignent leur
densité maximale chez les embryons, mais diminuent en nombre à mesure que les CNE des
branchies apparaissent. L'exposition d’embryons et de larves à une hypoxie chronique empêche
la perte des CNE de la peau, décale le pic de ventilation basale à un stade ultérieur de
développement, et induit une réponse hyperventilatoire à l'hypoxie aiguë. L'exposition chronique
à une hyperoxie résulte en une diminution rapide du nombre de CNE de la peau, déplace le pic
de ventilation vers des stades plus précoces de développement, et élimine la réponse à l'hypoxie
aiguë. L'administration de la neurotoxine 6-hydroxydopamine provoque la dégradation des
terminaisons nerveuses en contact avec les CNE de la peau et réduit à la fois la fréquence de
ventilation basale et la réponse hypoxique ventilatoire. Cette étude suggère un rôle pour les CNE
de la peau en tant que chémorécepteurs d’O2 extrabranchiaux chez le poisson-zèbre au cours du
développement.
iv
ACKNOWLEDGEMENTS
First I would like to thank my thesis supervisor Dr. Michael Jonz for giving me the
opportunity to continue working in his lab and to complete my project. Starting off as an
undergraduate student in your lab, and being able to continue my project as a graduate student
has been such a wonderful experience and you have continually inspired and encouraged me
throughout this entire process. Thank you for everything!
To my advisory committee, Drs. Marc Ekker and Steve Perry, thank you for taking the time to
be on my committee and providing me with valuable input. All of your questions and
suggestions helped me to make my project that much better.
To all the past and present Jonz lab members, all of you have helped me along my journey. I
especially would like to thank SA and PZ for your constant encouragement, kind words, and
introducing me to some great music (although I still can’t stand country). I wish you both the
best of luck during the rest of your studies. I also need to thank all of my close friends who have
provided me with wonderful support over my entire university career. VC, CB, BM, KM and
DP: you all have contributed to the person I am today, and I am proud to call you my friends. To
CR, who has always encouraged me to do my best and stay positive, thank you!
Finally I need to thank my family for encouraging me to pursue my dreams and to never give
up. To my mother and late father, thank you for instilling in me the importance of education, for
your unfaltering love and great care packages. To my sister, who has been a great roommate
here in Ottawa and a constant source of inspiration as she continues her journey towards a PhD,
and to my brother and sister-in-law for giving me encouragement to stay in school and for
always making me laugh. Also to my zio and zia here in Ottawa for the great Italian home
cooked meals.
v
TABLE OF CONTENTS
Abstract............................................................................................................................................ii
Résumé...........................................................................................................................................iii
Acknowledgements.........................................................................................................................iv
List of Tables and Figures.............................................................................................................vii
List of Abbreviations....................................................................................................................viii
1. Literature Review
1.1 Inroduction............................................................................................................................1
1.2 Oxygen sensing and respiratory chemoreceptors.................................................................2
1.2.1 Proposed sensors, mechanism and the hypoxic ventilatory response...........................3
1.2.2 The carotid body and adrenomedullary chromaffin cells.............................................5
1.2.3 Gill neuroepithelial cells.............................................................................................. 7
1.3 Ontogenesis of oxygen sensing in zebrafish.........................................................................8
1.4 Extrabranchial O2 chemoreceptors.....................................................................................10
1.5 Skin.....................................................................................................................................12
1.5.1 Innervation..................................................................................................................13
1.5.2 Surface chemoreceptors..............................................................................................13
1.6 Hypotheses and objectives..................................................................................................15
2. Materials and Methods
2.1 Animals...............................................................................................................................16
2.2 Immunohistochemistry.......................................................................................................16
2.3 Acclimation to hypoxia and hyperoxia...............................................................................19
2.4 Exposure to neurotoxin, 6-hydroxydopamine (6-OHDA).................................................19
vi
2.5 Ventilation frequency measurements..................................................................................21
2.6 Morphometric analysis........................................................................................................22
2.7 Statistics..............................................................................................................................23
3. Results
3.1 General features of skin neuroepithelial cells (NECs)........................................................24
3.2 Development of skin NECs and effects of water PO2........................................................29
3.3 Development of the hypoxic response................................................................................37
3.4 Innervation of skin NECs by catecholaminergic nerve fibres............................................38
4. Discussion..................................................................................................................................48
4.1 Skin NECs as oxygen chemoreceptors...............................................................................48
4.2 The hypoxic ventilatory response.......................................................................................55
5. Conclusions and future directions..............................................................................................58
6. References..................................................................................................................................60
Appendix I.....................................................................................................................................66
vii
LIST OF TABLES AND FIGURES
Table 1. Primary and secondary antibodies used for immunohistochemistry.
Figure 1. Distribution of serotonergic (5-HT)-positive skin NECs of larval and adult zebrafish
as labelled with an antibody against 5-HT.
Figure 2. Innervation of skin NECs with zn-12-positive nerve fibres.
Figure 3. Skin-NECs and their innervation in embryonic zebrafish as labelled with 5-HT and
zn-12 antibodies.
Figure 4. The number and density (mm-2
) of NECs found on the skin of developing zebrafish
(3-9 d.p.f.) is dependent on developmental stage and water PO2.
Figure 5. The size of skin NECs is affected by chronic hyperoxia but not chronic hypoxia in 3-9
d.p.f. zebrafish.
Figure 6. Development of behavioural responses to hypoxia in zebrafish from 1-9 d.p.f.
Figure 7. Chronic exposure of zebrafish to hypoxia and hyperoxia caused a hypoventilatory
response to hypoxia and a shift in peak basal ventilation frequency.
Figure 8. 6-hydroxydopamine (6-OHDA) exposure leads to neuronal loss in the skin of
developing zebrafish and loss of contact between skin NECs and zn-12 positive nerve
fibres.
Figure 9. 6-OHDA treatment reduced basal ventilation frequency and eliminated response to
acute hypoxia.
Figure 10. Chronology of the development of ventilatory events and structures in zebrafish.
Appendix I. Length and heart rate of developing zebrafish during chronic hypoxia exposure.
viii
LIST OF ABBREVIATIONS
5-HT, 5-hydroxytryptamine or serotonin
6-OHDA, 6-hydroxydopamine
AMCs, adrenomedullary chromaffin cells
ANOVA, analysis of variance
ASR, aquatic surface respiration
CB, carotid body
CNS, central nervous system
d.p.f., days post-fertilization
FITC, fluorescein isothiocyanate
h.p.f., hours post-fertilization
NEBs, neuroepithelial bodies
NECs, neuroepithelial cells
PBS, phosphate buffered solution
PO2, partial pressure of oxygen
SCCs, solitary chemosensory cells
SEM, standard error of mean
TBs, taste buds
zn-12, zebrafish specific neuronal antibody
1
1. LITERATURE REVIEW
1.1 Introduction
The focus of this thesis has been to determine if skin neuroepithelial cells (NECs) of
developing zebrafish are oxygen sensitive and if they may play a role in mediating the
ventilatory response to low O2 (hypoxia). This section will review the pertinent literature on
respiratory chemoreceptors in vertebrates and give examples of these cells in both the
mammalian and fish models. The development of O2 sensing in zebrafish embryos and larvae is
then outlined, and the importance of extrabranchial O2 chemoreceptors in ventilatory control is
discussed. Finally, the skin is then discussed as playing an important respiratory role during
development, and details about its innervation and known populations of chemoreceptors are
included to provide context of this organ as an established site of chemosensing.
The zebrafish, Danio rerio, is an excellent model that has been used for explanations
pertaining to many elements such as vertebrate development and gene expression. The ease of
breeding and maintenance in the laboratory make this an attractive animal model to use in
developmental studies. Zebrafish embryos and larvae are amenable to imaging, and much is
known about the ontogenesis of oxygen sensing of this species. Additionally, this animal
represents the only non-mammalian species for which O2 chemoreceptors have been
characterized in larvae and adults (Jonz and Nurse, 2006). Through the current study, the
zebrafish will be further developed as a model of respiratory research and oxygen sensing. As
well, this study will improve the overall understanding of the development of oxygen sensing in
vertebrates.
2
1.2 Oxygen sensing and respiratory chemoreceptors
The ability to sense and respond to levels of oxygen is vital for the survival of many
organisms, including vertebrates. Respiratory chemoreceptors are specialized cells that detect
changes in the level of environmental or arterial O2 or CO2 and initiate compensatory changes in
heart rate, ventilation and release of catecholamines (Burlson and Milsom, 2003; Lopez-Barneo
et al., 2008; Perry et al., 2009). Because of their functionality, respiratory chemoreceptors have
been morphologically identified according to the following general criteria: access to a sampling
environment (external or arterial), retention of neurotransmitter-containing synaptic vesicles for
neurosecretion, and sensory innervation (Gonzàlez et al., 1994; Jonz and Nurse, 2005).
Respiratory chemoreceptors are well-described in mammals, such as the pulmonary
neuroepithelial bodies (NEBs) of the lungs, carotid body (CB) type I cells and neonatal
adrenomedullary chromaffin cells (AMCs; Cutz and Jackson, 1999; Lopez-Barneo et al., 2008;
Nurse et al., 2009). These cells are biochemically and morphologically similar to the NECs of
the fish gill, a population of cells first described by Dunel-Erb et al. (1982) that were suspected
of having O2 chemosensory capabilities. More recently, NECs of the adult zebrafish gill were
shown to display morphological properties (i.e. retention of serotonin, synaptic vesicles and
innervation) and physiological responses to hypoxia, such as ion channel inhibition and
membrane depolarization, that are typical of O2 chemoreceptors (Jonz and Nurse, 2003; Jonz et
al., 2004). Furthermore, exposure to chronic hypoxia induces hypertrophy, and proliferation in
NECs (Jonz et al., 2004), which are typically observed in mammalian O2 chemoreceptors (Nurse
and Vollmer, 1997; Wang and Bisgard, 2002).
3
1.2.1 Proposed sensors, mechanisms and the hypoxic ventilatory response
Oxygen sensing is very important in order for vertebrates to maintain homeostasis in
environments which experience oxygen fluxes. Fish are regularly affected by changes in oxygen
saturation in water. In fact, hypoxia is a natural component of many freshwater habitats (Diaz
and Breitburg, 2009). Depending on temperature and salinity, water contains 20-40 times less
O2 by volume and O2 diffuses about 10000 times more slowly through water than air (Diaz and
Breitburg, 2009). In order to respond to O2 flux, there must be sensors in place to recognize this
change, a mechanism to transmit this information to the central nervous system (CNS) and a
compensatory response can then occur.
There has been much research on the subject of how oxygen is sensed in respiratory
chemoreceptors. Thus, two hypotheses have emerged for the site of the molecular O2 sensor: the
mitochondrial model and the membrane model (Lahiri et al., 2006; López-Barneo et al., 2008).
The mitochondrion uses oxygen as the final electron acceptor, and, similarly to hypoxia,
inhibitors of the electron transport chain increase the afferent activity of the CB sinus nerve
(Mills et al., 1972). However, this model has lost much support due to other findings. For
example, inhibitors of the mitochondrial electron transport chain do not block the hypoxia-
induced release of transmitter (Ortega-Sáenz et al., 2003). Therefore, it is generally accepted
that mitochondria do not act as O2 sensors but changes in mitochondrial function may contribute
to the hypoxia-induced rise of cytosolic Ca2+
concentration necessary to trigger respiratory
chemoreceptor secretion, although the details for this action are not known (Lahiri et al., 2006).
The membrane hypothesis, originally proposed for the CB (López-Barneo et al., 1988),
predicts that O2 is sensed at the plasma membrane. While a membrane-delimited O2 sensor is
dependent on species and is controversial as a general mechanism, there is at least a consensus in
4
the literature that transduction of the hypoxic stimulus is mediated by inhibition of plasma
membrane K+ channels, and this produces the necessary depolarization for neurosecretion (Perry
et al., 2009). The current view of O2 sensing in vertebrate chemoreceptors suggests that a variety
of molecular sensors may be involved, and that regulation of plasma membrane K+ channels may
be linked to mitochondrial respiration (Perry et al., 2009).
Regardless of the molecular mechanism of O2 sensing in chemoreceptors, an accepted view of
the downstream effects of hypoxia on chemoreceptors has been established for both mammals
and fish, in which reduced PO2 leads to the following sequence of events: (1) inhibition (closure)
of membrane-bound K+ channels; (2) membrane depolarization; (3) activation (opening) of
voltage-gated Ca2+
channels; (4) increase in cytosolic Ca2+
concentration; (5) neurotransmitter
release from cytoplasmic synaptic vesicles; and (6) activation of afferent nerve fibres (López-
Barneo et al., 2008; Perry et al., 2009).
Once hypoxia is sensed by the respiratory chemoreceptors and the signal is transmitted, a
number of physiological, anatomical and behavioural adaptations can occur. Manners in which
fish can adapt to hypoxia are by reducing metabolic rates, accessing highly oxygenated
environments, aquatic surface respiration (ASR) and hyperventilation (Diaz et al., 2009; Perry et
al., 2009). The hypoxic ventilatory response (i.e. hyperventilation) has been identified as the
most important physiological response of fish upon exposure to hypoxia (Perry et al., 2009).
The majority of sole water-breathing fish respond to acute hypoxia by increasing the volume of
water ventilated (see Table 5.1, Perry et al., 2009). In adult zebrafish, acute hypoxia exposure
led to an increase in breathing frequency and no change in breathing amplitude (Vulesevic et al.,
2006).
5
The hypoxic ventilatory response serves to maintain oxygen uptake when oxygen availability
declines. Gas transfer across the gills involves the net direction of oxygen inward, as oxygen is
extracted from the water and travels into the bloodstream and the net direction of carbon dioxide
to be outward, as CO2 is excreted by the blood and goes into the water (Perry and Gilmour,
2002). Hyperventilation leads to the increase of PO2 and decrease of PCO2 by reducing the
carbon dioxide concentration in the blood to below its normal level, as more CO2 is being
excreted than is being produced, leading to the alteration of the blood-to-water diffusion
gradient. Through the increase of water flow across the gills, the inspired-expired PO2 difference
decreases and serves to minimize the extent of the reduction in arterial PO2 (Perry et al., 2009).
1.2.2 The carotid body and adrenomedullary chromaffin cells
Since CB type I cells are relatively insensitive to hypoxia during the first few weeks of life,
AMCs have been identified as the oxygen chemoreceptors during the perinatal period in
mammals (Thompson et al., 1997). These cells, located in the adrenal glands, are important for
survival of the newborn during birth as a high level of hypoxic stress is experienced at this time
(López-Barneo et al., 2010). As demonstrated in rats, AMCs possess a developmentally
regulated O2 sensing mechanism (Thompson et al., 1997). AMCs of neonatal rats responded to
hypoxia through an increase in cytosolic Ca2+
concentration, resulting in membrane
depolarization as a result of suppressed voltage-gated K+ channels, and a release of
catecholamines (Thompson et al., 1997; Rico et al., 2005). About 2-3 weeks postnatal, the
sensitivity to hypoxia is lost following pre-ganglionic cholinergic innervation (Thompson et al.,
1997; Nurse et al., 2009). Coincidentally at this time, CB type I cells mature and become
sensitive to hypoxia (Slotkin and Seidler, 1988).
6
The CB is a small, paired, neural-crest derived organ located at the carotid bifurcation of adult
mammals and is the principal acute O2-sensing organ (López-Barneo et al., 2010). Upon
exposure to chronic hypoxia, the CB undergoes a number of morphological and physiological
changes: enlargement and mitosis of the type I cells, growth of new blood vessels and
vasodilation, all leading to CB hypertrophy (Stea et al., 1992; Mills and Nurse, 1993; Wang and
Bisgard, 2002). These changes serve to increase the sensitivity of the CB chemoreceptors to
hypoxia in adults (Barnard et al., 1987; Bisgard, 2000). However, it has been demonstrated that
humans native to high altitudes and developing neonates exposed to chronic hypoxia may have a
blunted ventilatory response to acute hypoxia due to a decreased O2-sensitivity of the carotid
body (Bisgard, 2000; Lahiri et al., 2000). Similarly, chronic hyperoxia exposure during
development alters ventilation in adult life due to blunted afferent inputs from the carotid body
and neurons in the carotid sinus nerve (Bisgard et al., 2005).
The O2-sensitive type I cells of the CB are innervated by sensory nerve fibres of the carotid
sinus nerve and by postganglionic sympathetic nerve fibres from the superior cervical ganglion
(Gonzalez et al., 1994; Wang and Bisgard, 2002). The CB type I cells have been shown to
exhibit catecholaminergic efferent and afferent innervation (Katz et al., 1983; Finley et al.,
1992). It has also been demonstrated that carotid body afferent fibres may release
neurotransmitter (dopamine) onto type I cells and is supported by observations that these fibres
contain synaptic vesicles and presynaptic membrane specializations (McDonald and Mitchell,
1975; Smith and Mills, 1976).
7
1.2.3 Gill neuroepithelial cells
The NECs found on the gills of zebrafish have been recently identified as bimodal O2 and
CO2 chemoreceptors (Jonz et al., 2004; Qin et al., 2010). A diverse number of NECs have been
described in the zebrafish gill, but the serotonergic NECs of the gill filament have been identified
as the O2-sensitive cells due to their exposure to the external environment, synaptic vesicle
retention, innervation and electrophysiological recordings during hypoxic exposure (Jonz and
Nurse, 2003; Jonz et al., 2004).
Similarly to the CB, chronic hypoxia induces morphological change to gill NECs. After in
vivo exposure of zebrafish to chronic hypoxia, the gill NECs increased in size by approximately
15% compared to controls (Jonz et al., 2004). More NECs had neuron-like processes after
chronic hypoxia, and these extensions were longer in exposed fish versus controls (Jonz et al.,
2004). No change in density of serotonergic gill NECs has been reported after chronic hypoxia
exposure, although the density of non-serotonergic NECs did increase under these conditions
(Jonz et al., 2004; Vulesevic et al., 2006). Upon exposure of fish to chronic hyperoxia, the
density of serotonergic gill NECs was significantly reduced (Vulesevic et al., 2005). Chronic
exposure of adult zebrafish to hypoxia had no effect on their ventilatory response to acute
hypoxia, whereas chronic exposure to hyperoxia significantly blunted breathing frequency upon
acute hypoxia exposure (Vulesevic et al., 2005). Zebrafish reared under hypoxic conditions
showed no change in their response to acute hypoxia as adults, whereas zebrafish reared under
hyperoxic conditions showed a blunted response to acute hypoxia as adults (Vulesevic and Perry,
2006).
In teleost fish, there are generally at least three types of nerve fibres with which gill NECs
establish synapses. The first type is characterized by catecholaminergic nerve endings,
8
suggesting that NECs are modulated by spinal autonomic nerves (Bailly et al., 1992; Jonz and
Zaccone, 2009). The second type is made of serotonergic neurons, which are not degenerated
after treatment with 5- or 6-hydroxydopamine (5- or 6-OHDA; Bailly et al., 1992). The third
type is nitrergic nerve fibres that may be of cranial origin (Zaccone et al., 2003). The NECs may
also receive efferent innervation, as many nerve fibres associated with gill NECs contain
synaptic vesicles representative of presynaptic nerve endings (Jonz and Nurse, 2003; Jonz and
Zaccone, 2009).
1.3 Ontogenesis of oxygen sensing in zebrafish
Zebrafish go through many stages of development before gill NECs and associated sensory
pathways, which are necessary for oxygen sensing in adults, are functional. Prior to the
development of gill NECs, zebrafish are able to sense oxygen levels (Jonz and Nurse, 2005).
These pathways, however, are not fully developed until 7 d.p.f. (Jonz and Nurse, 2006). In order
to enable the developing zebrafish to respond to hypoxic challenges, there must be alternative
sites of O2 chemoreceptors.
When subjected to 24 hours of anoxia, zebrafish embryos 25 h.p.f. and younger survived by
entering a state of metabolic suppression where all cell division, cardiac function, and movement
ceased (Padilla and Roth, 2001). Upon re-oxygenation, embryos were able to develop normally
and produce offspring that were indistinguishable from fish raised under normal conditions. This
type of anoxia tolerance has been observed in other teleost fish such as the annual killifish,
Austrofundulus limnaeus (Podrabsky et al., 2007) and several mammalian species (Renfree and
Shaw, 2000; Kamei et al., 2011). Tolerance to anoxia can be seen as a protective measure to
9
enhance survival, increase reproductive fitness or as an evolutionary adaptation to natural
environmental fluctuations (Padilla and Roth, 2001; Podrabsky et al., 2007).
By 2 d.p.f., zebrafish embryos lose their ability to tolerate anoxia, and mortality is induced
(Padilla and Roth, 2001). The decrease in duration of anoxic viability has been shown to
correlate with the rate of anoxic lactate accumulation (Mendelsohn et al., 2008). During the first
few days of development, there is a significant increase in O2 consumption (Barrionuevo and
Burggren, 1999), which may be related to this decreased anoxia tolerance. Also at this stage of
development, zebrafish embryos responded to hypoxia by increasing pectoral fin and whole body
movements (Jonz and Nurse, 2005), improving the diffusion gradient for respiratory gas
exchange across the skin (Rombough, 1988). Whole body and pectoral fin movements have
been shown to be coordinated with gill ventilation, and can be used to identify a response to
hypoxia before the hyperventilatory response develops (Jonz and Nurse, 2005). At 2 d.p.f., there
appears to be no morphological basis for gas exchange to occur across the gills, because gill
NECs and respiratory lamellae are not present (Rombough 2002; Jonz and Nurse, 2006).
Zebrafish predominately respire cutaneously up to 14 d.p.f., and respiratory gas exchange
eventually shifts to the gills (Rombough, 2002).
By 3 d.p.f. gill development begins with the appearance of filament primordia, innervation of
gill arches and gill arch NECs, and the gills are ventilated (Jonz and Nurse, 2005; Higashijima et
al., 2000; Rombough, 2002). Despite the apparent lack of gas exchange across the gills, the
hyperventilatory response to hypoxia also begins at this stage (Jonz and Nurse, 2005). At 5 d.p.f.
the gill NECs, which have been demonstrated to be responsible for O2-chemorecepton in adults
(Jonz and Nurse, 2006), are present and their innervation begins and is completed by 7 d.p.f.
(Jonz and Nurse, 2005). As the gills mature, the contribution of cutaneous gas exchange
10
decreases, and the gills become the dominant site for respiration by 21 d.p.f. (Rombough, 2002).
Since the NECs responsible for O2 sensing are neither innervated nor functional until 7 d.p.f.,
they are unlikely to serve as oxygen chemoreceptors until this time. This phenomenon suggests
that embryonic and larval O2 sensing must occur via chemoreceptors that occupy extrabranchial
sites.
1.4 Extrabranchial O2 chemoreceptors
The gill NECs of fish have been well documented to be the oxygen-sensitive cells of the adult
teleost fish (Dunel-Erb et al., 1982; Bailly et al., 1992; Jonz et al., 2004; Jonz and Nurse, 2006).
However, sensitivity to hypoxia appears early in zebrafish development, well before the gill
NECs are functional (see section 1.3), indicating that extrabranchial sites of O2 sensing exist
during these early stages, similarly to mammalian foetuses (Schwerte, 2009). Several
suggestions have been made as to where extrabranchial chemoreceptors are located in adult fish,
such as the pseudobranch, the central nervous system (brain), and orobranchial cavity (Perry et
al., 2009; Schwerte, 2009). One study has proposed that extrabranchial O2 sensing is a
developmental phenomenon, and may occur in the skin (Jonz and Nurse, 2006).
The pseudobranch is a reduced, gill-like organ that resides within the cranial portion of the
opercular epithelium, and is not exposed to the external environment. It has been proposed to be
involved in O2 sensing on the basis of its morphology and responsiveness to hypoxic perfusate
(Laurent and Rouzeau, 1972). However, additional studies on the pseudobranch were unable to
elicit the same response characteristic, and removal of the pseudobranch did not change the
ventilatory response to hypoxia (Parry and Holliday, 1960; Bamford, 1974). As well, many
species that do not have a pseudobranch (e.g. channel catfish) still possess robust responses to
11
hypoxia (Perry et al., 2009). Therefore, the role of the pseudobranch in oxygen sensing is
inconclusive.
The central nervous system (brain) has been implicated as a site of oxygen chemosensing.
Experiments demonstrated that infusion of hypoxic blood into the dorsal aorta of a normoxic sea
raven elicited a hyperventilatory response (Saunders and Sutterlin, 1971). Conversely, a study
performed on the neotropical fish Colossoma macropomum (tambaqui) demonstrated that the
brain is not the site of the oxygen chemoreceptors, as superfusion of hypoxic saline did not elicit
the hyperventilatory response (Milsom et al., 2002). Therefore, central O2 sensing may be
species-specific. It has been demonstrated in adult fish that increased ventilation frequency in
response to hypoxia is reliant on central ionotropic glutamate receptors (NMDA receptors;
Turesson and Sundin, 2003). In zebrafish, these receptors are necessary for mediating the
hypoxic ventilatory response beginning at 8 d.p.f. and this response is completely dependent on
NMDA receptors by 13 d.p.f. (Turesson et al., 2006). This study also found that the NMDA
receptor subunits NMDAR1 and NMDAR2A are not present in fish as young as 3 d.p.f.,
suggesting that the hyperventilatory response is not dependent on NMDA receptors during early
developmental stages, and other mechanisms may be responsible (Turesson et al., 2006).
The orobranchial cavity has been suggested as a site of extrabranchial O2 chemoreceptors. In
a neotropical fish, the tambaqui (Colossoma macropomum), complete denervation of the gill did
not eliminate the hypoxic response (Milsom et al., 2002). This same study provided evidence to
support the orobranchial cavity, innervated by cranial nerves V and VII, as possessing O2-
sensitive chemoreceptors. It was shown that the chemoreceptors in this area were involved in
reflex increases in breathing amplitude (Milsom et al., 2002). Therefore, it remains a possibility
that such chemoreceptors of the buccal cavity may be present in zebrafish.
12
To date, the only tissue that has been considered as a site for extrabranchial O2
chemoreception in developing fish is the skin (Jonz and Nurse, 2006). There is evidence of
many types of chemosensory cells in the skin which direct information to the brain (see section
1.5), therefore it is reasonable to propose that chemoreceptors which detect O2 may also be
present in the skin. The skin has been identified as an accessory respiratory surface, and the
upper layer of the dermis is most strongly vascularised (Le Guellec et al., 2004). Only one study
has described the presence of serotonergic (5-HT) positive cells on the skin of 3 d.p.f. zebrafish
that resemble O2 chemoreceptors of the gills (Jonz and Nurse, 2006). These cells are distributed
all over the surface of the zebrafish and appear to be associated with nerve fibres. A similar
population of cells has also been observed in Xenopus larvae (Jonz and Nurse, 2006).
1.5 Skin
The skin of fish provides a protective barrier between the animal and its aquatic environment.
Several important functions are encompassed by the skin including maintaining body shape,
protecting the fish from shocks and various attacks, improving hydrodynamics and housing
sensory functions that are essential to survival (Le Guellec et al., 2004). The skin can be divided
into three compartments: the epidermis, dermis and hypodermis (Le Guellec et al., 2004). The
epidermis is composed of several regions, and is penetrated by nerve fibres which are either free
nerve endings or associated with epidermal sensory cells (Whitear, 1971). Before 24 h.p.f., the
skin of zebrafish is comprised of only a two cell layer (about 4 μm thick) of epithelial cells
which surround the body, and the skin remains thin (>10 μm) throughout embryonic and larval
stages (Le Guellec et al., 2004). Even during these early stages, the skin remains a highly
functioning organ as it has been implicated to be the main site of gas exchange, and contains
13
numerous chemosensory cells necessary for survival (Kotrschal et al., 1997; Rombough, 2002;
Hansen et al., 2002; Døving and Kasumyan, 2008).
1.5.1 Innervation
The head epidermis of teleosts receives dual sensory innervation by trigeminal and facial
fibres, while the trunk receives facial and spinal innervation (Kotrschal et al., 1997). Single areas
of skin have been found to be innervated by multiple nerve fibres, similarly to mammalian skin
(Whitear, 1952). As described by Whitear (1971), there are four sensory components of fish skin
innervation: (1) somatic, which innervates the organs of the lateral line system (neuromasts); (2)
visceral, which innervates taste buds; (3) general cutaneous, which innervates the outer skin
surface with receptor fibres for tactile, temperature and pain senses, and also produce free nerve
endings; and (4) general visceral, which produces free nerve endings in the oral and pharyngeal
epithelia.
1.5.2 Surface chemoreceptors
There are three chemosensory systems in fish: (1) the olfactory system, the sense of smell,
can detect substances of minute quantities and is localized to the snout in a peripheral organ; (2)
the gustatory system, the sense of taste, detects chemicals at a higher concentration than the
olfactory system. The taste receptor cells (taste buds; TBs) can be found in the oral cavity and
over the entire skin surface; (3) the common chemical sense, the function of which has not been
determined, is presented by free nerve endings, and solitary chemosensory cells (SCCs) which
are located on the entire skin surface (Døving and Kasumyan, 2008). Only the TBs and SCCs
14
will be discussed here, as they are distributed over the entire skin surface, similarly to the
previously identified 5-HT cells (see section 1.4).
Mature fish TBs are ovoid in shape, about 80-100 µm high and 40-60 µm wide, and found in
the epithelium, sitting on a small dermal papilla (Hansen and Reutter, 2004). They can be found
in many areas of the zebrafish, including the epithelia of the lips and oropharyngeal cavity, and
also on the skin of the barbels, head and whole body surface (Hansen et al., 2002). Innervation
of TBs is derived from the facial (VII), glossopharyngeal (IX), or vagal (X) cranial nerves
(Døving and Kasumyan, 2008). TB primordia do not appear on the zebrafish until 3-4 days post-
fertilization (d.p.f.)., and are isolated to the skin of the lips and gill arches. TBs of the mouth and
oropharyngeal cavity appear one day later (4-5 d.p.f.) and around 12 d.p.f. is when the TBs
appear on the surface of the head (Hansen et al., 2002).
All groups of fish have been shown to possess SCCs, which are solitary, spindle-shaped, and
have single or multiple microvilli (Whitear, 1992; Døving and Kasumyan, 2008). SCCs appear
in the zebrafish after hatching, approximately 3 d.p.f., and continue to increase in density after
metamorphosis (25 d.p.f.), remaining relatively constant thereafter (Kotrschal et al., 1997). They
are found all over the surface of the fish, with higher densities occurring in the head, rather than
the trunk (Kotrschal et al., 1997). Innervation of SCCs depends on location: the head and
oropharyngeal SCCs can receive facial or spinal innervation, while SCCs on the body surface
may receive innervation from a recurrent facial nerve (Whitear, 1992). The role of SCCs has not
been elucidated, however these cells have been proposed to play a role in predator avoidance or
to be the evolutionary precursors of TBs (Kotrschal, 1991).
15
1.6 Hypotheses and objectives
The question which drove this research was: are skin NECs oxygen sensitive chemoreceptors
during early development in zebrafish? To address this question, a number of hypotheses were
tested:
1) It was hypothesized that if skin NECs are oxygen chemoreceptors in embryos and larvae,
they will share similar morphological characteristics with oxygen chemoreceptors of the gills in
adults, such as innervation by sensory nerve fibres. It is also predicted that skin NECs will be
observed primarily during the embryonic and early larval stages, before gill NECs develop.
These hypotheses will be addressed using immunohistochemistry and fluorescence microscopy.
2) Chronic exposure to hypoxia or hyperoxia early in development will induce morphological
changes in skin NECs, such as changes in number, density and size, which have been observed
in O2 chemoreceptors of both fish and mammals. Furthermore, since embryos and early larvae
exhibit behavioural responses to changes in water PO2, the above acclimation to chronic hypoxia
or hyperoxia will also lead to changes in the way the larvae respond to acute hypoxia. These
hypotheses will be addressed using morphometric analysis of immunolabelled skin NECs and
behavioural assays.
3) Innervation of skin NECs is required for the preservation of the embryonic and larval
response to acute hypoxia. This hypothesis will be tested by pre-exposing developing zebrafish
to a sublethal concentration of a neurotoxin that selectively degrades catecholaminergic nerve
endings. These embryos and larvae will then be subjected to behavioural assays to determine
changes in the acute response to hypoxia.
16
2. MATERIALS AND METHODS
2.1 Animals
Animals used in this study were developing zebrafish (Danio rerio) ranging from 24 hours
post-fertilization (h.p.f.) to 9 days post-fertilization (d.p.f.). Adult zebrafish, used for breeding,
were obtained from a commercial supplier (MIRDO, Montreal, Canada) and transported to the
University of Ottawa Aquatic Care Facility and maintained in 9 litre acrylic tanks supplied with
aerated, dechloraminated City of Ottawa tap water at 28°C. Fish were maintained on a constant
10:14 light-dark photoperiod. Embryos were obtained using the standard zebrafish breeding
technique (Westerfield, 2000). Briefly, plastic traps were placed in tanks with mixed males and
females. Embryos were collected through a fine mesh the following morning and placed in
embryo medium in an incubator equilibrated with room air and set at 28.5°C. Embryo medium
(E3) contained the following chemicals (in mM): NaCl, 5; KCl, 0.17; CaCl2·2H2O, 0.33;
MgSO4·7H2O, 0.33; 0.0001% (w/v) methylene blue; pH 7.78. After 2 days, embryos were
transferred to 2 litre plastic tanks filled with dechlorinated water maintained at 28.5°C in a water
bath. Embryos and larvae were not fed during experiments. Care and handling of animals was
done according to guidelines set out by the University of Ottawa Veterinary Service and carried
out in accordance with those of the Canadian Council on Animal Care (CCAC).
2.2 Immunohistochemistry
Once the larvae reached the desired chronological age, their chorion was removed (if
necessary; Westerfield, 2000). Tissue was fixed in a cold phosphate buffered solution (PBS)
containing 4% para-formaldehyde and kept at 4°C overnight. PBS contained the following
chemicals (in mM): NaCl, 137; Na2HPO4, 15.2; KCl, 2.7; KH2PO4, 1.5; pH 7.7-7.8 (Bradford et
17
al., 1994; Jonz and Nurse, 2003). The larvae were then rinsed in cold PBS. The tissue was
placed overnight at 4°C in a permeabilizing solution, containing 2% Triton X-100 in PBS. This
concentration of Triton X-100 was chosen because preliminary experiments using a lower
concentration (0.5%) resulted in poor labelling. The higher concentration improved labelling
without any adverse effects (i.e. no non-specific labelling or background, etc.).
For experiments in which skin was isolated for immunohistochemistry, pieces of skin were
collected from 3 month old adult zebrafish after a sharp blow to the head and decapitation. The
tissue was pinned down into a Sylgard (184 Silicone Elastomer Kit, Dow Corning Corporation,
Midland, MI) coated petri dish and kept in PBS. Under a dissecting microscope (Leica MZ6),
skin was separated from underlying connective tissue using fine forceps and kept on ice until
fixation.
Morphological characteristics of serotonergic skin cells were identified using an array of
commercially available antibodies (Table 1) and previously developed procedures (Jonz and
Nurse, 2003). Identification of gill NECs and serotonergic skin cells have previously been
performed using anti-5-HT, diluted 1:250 (Jonz and Nurse, 2003; Coccimiglio, 2008).
Innervation of the serotonergic skin cells was detected using the primary antibody zn-12
(neuron-specific surface antigen; Trevarrow et al., 1990). Monoclonal zn-12 was raised in
mouse against membrane fractions from adult zebrafish CNS and recognizes an HNK-1-like
(human natural killer-1, a carbohydrate expressed on certain neural cell adhesion molecules)
epitope (manufacturer specifications). Western blot analysis has indicated that both zn-12 and
HNK-1 antibodies label similar bands ranging in molecular weight from 60-248 kDa (see
Metcalfe et al., 1990). Permeabilized tissue was bathed in the desired primary antibodies for 24
hours at 4°C and then exposed to the appropriate secondary antibody after rinsing with
18
permeabilizing solution. Positive immunofluorescence was identified using secondary
antibodies conjugated with fluorophores (e.g. FITC and Alexa 594). The tissue was exposed to
the secondary antibodies for 50-60 minutes in a dark, room temperature environment. Prior to
mounting on slides, the tissue was rinsed with PBS, placed in a mounting medium (Vectashield,
Vector Laboratories Inc., Birlingame, CA) and then viewed using a confocal microscope
(Fluoview 200, Olympus), or microscope equipped with epifluorescence (Axiophot (Zeiss or
Nikon). The confocal microscope was equipped with argon (Ar) and krypton (Kr) lasers with
peak outputs of 488 nm and 568 nm respectively. A photomultiplier tube detected images and
photographs were captured using confocal graphics software (Fluoview 2.1, Olympus). Tissue
was scanned using the appropriate laser for each fluorophore. When conventional
epifluorescence microscopy was used, images were collected with Northern Eclipse software
(Empix Imaging Inc., Mississauga, ON, Canada). All image processing and manipulation was
done with ImageJ software (v. 1.42q; National Institutes of Health).
Table 1. Primary and Secondary Antibodies Used for Immunohistochemistry.
______________________________________________________________________________ Antisera Dilution Antigen Host Source Secondary
Primary
5-HT 1:250 serotonin rabbit Sigma FITC
zn-122,4
1:100 neuron, surface mouse DSHB3 Alexa 594
Secondary1
Alexa 594 1:100 mouse (IgG) goat Molecular Probes
FITC5 1:50 rabbit (IgG) goat Jackson Laboratories
______________________________________________________________________________
1Secondary antisera was conjugated with a fluorescent marker.
2Monoclonal antibody.
3Developmental Studies Hybridoma Bank, University of Iowa.
4Zebrafish derived antibody.
5Fluorescein isothiocyanate.
19
2.3 Acclimation to hypoxia and hyperoxia
To study the impact of water PO2 on zebrafish development, embryos and larvae were
acclimated to one of four different levels of water PO2: hyperoxia (300 Torr), normoxia or
control (150 Torr), mild hypoxia (80 Torr), or severe hypoxia (30 Torr). For mild hypoxia
acclimation experiments, fresh embryos were placed in an incubator equipped with an O2 sensor
and pumped with a mixture of air and nitrogen gas to gradually acclimate the embryos to the
desired level of hypoxia. Embryo medium was replaced daily to minimize mortality. When
larvae reached 2 d.p.f. they were placed in dechlorinated water bubbled with compressed air and
nitrogen gas using a Pegas 4000 gas mixer (Columbus Instruments, Columbus, OH) to maintain
the required level of hypoxia. An oxygen meter (YSI model 55, YSI Inc., Yellow Springs, OH)
was used to test the level of dissolved oxygen in the water every day. Control embryos were
held in normoxia in which water was bubbled with compressed air only.
For more severe hypoxic treatment, embryos were housed in an incubator until 6 h.p.f. and
then placed in dechlorinated water and bubbled with the above combination of gases to achieve
the desired hypoxia level. The earlier transfer of embryos exposed to 30 Torr hypoxia promotes
survival (Vulesevic and Perry, 2006). Chronic hyperoxia exposures were performed in the same
manner as above, except water was bubbled with 100% oxygen to achieve the desired oxygen
level (300 Torr). Control embryos were also transferred to dechlorinated water at 6 h.p.f. and
were bubbled with compressed air only.
2.4 Exposure to neurotoxin, 6-hydroxydopamine (6-OHDA)
6-OHDA has been shown to selectively degrade catecholaminergic nerve endings in
mammalian models and the zebrafish (Blum et al., 2001; Parng et al., 2007). 6-OHDA, a
20
hydroxylated analog of the natural dopamine neurotransmitter, shows a high specificity to
catecholaminergic cells because of their specific uptake transporters (Blum et al., 2001).
Broadly speaking, 6-OHDA exerts its deleterious effects through inducing oxidative stress and
decreasing cellular adenosine triphosphate (ATP) levels. Cell death is said to occur by three
main mechanisms: 1) the creation of reactive oxygen species (ROS); 2) production of hydrogen
peroxide induced by monoamine oxidase activity; and 3) direct inhibition of the mitochondrial
respiratory chain (Blum et al., 2001).
For 6-OHDA (Sigma) treatment, previously determined procedures were employed (Parng et
al., 2007) with some modifications. 2 d.p.f. larvae from the same pool of embryos were split
into separate 50 ml beakers and administered 6-OHDA or left untreated (for negative control).
6-OHDA was applied directly to the fish water in 250 μM or 500 μM concentrations. Varicosity
distance and number were observed 3 and 5 days post-treatment (5 and 7 d.p.f.; see section 2.6).
Fish were tested in ventilation assays (see section 2.5) everyday from 1-5 days post-treatment (3-
7 d.p.f.). Negative controls were also tested at these same time points. Fish water containing 6-
OHDA was replaced after 3 days to minimize mortality.
Ventilation frequency measurements of fish acutely exposed to sodium cyanide (NaCN,
Sigma) were completed in the same manner as described in section 2.5, with some modifications.
Control and 6-OHDA treated fish were raised until 5 d.p.f. Normoxia measurements were taken
3 minutes after transfer using control solution. Fish were then exposed to a 1 mM solution of
NaCN dissolved in dechlorinated water for 1 minute before measurements were recorded.
21
2.5 Ventilation frequency measurements
Previously determined procedures for measuring ventilation frequency were employed (Jonz
and Nurse, 2005). Once the embryos (or larvae) achieved the desired stage of development (1
d.p.f. to 9 d.p.f.), the zebrafish was placed into a glass bottomed culture dish (MatTeck
Corporation, Ashland, MA) which has a 14 mm well, and continuously perfused at 4 ml/min.
Larvae ≥ 3 d.p.f. were lightly anaesthetized with 0.05 mg/ml MS222 (tricane methanesulfonate;
Syndel Laboratories, Vancouver BC) dissolved in dechlorinated water. Zebrafish were
transferred to the culture dish using a Pasteur pipette containing a small volume of dechlorinated
water. A piece of nylon mesh was placed over the bottom of the dish to confine the zebrafish to
the glass bottomed well. A moulded Sylgard ring was placed over the mesh to keep it in place
during perfusion. Observations were made using a dissecting microscope (Leica MZ6).
Behavioural response to hypoxia was determined by observing the rate of whole
body/pectoral fin movements in embryos aged 1 d.p.f. to 2 d.p.f., while movement of the
gills/operculae, and buccal pumping was monitored for older larvae (≥ 3 d.p.f.). Activity of gill
ventilation has been observed to be coordinated with pectoral fin and whole body movements in
larvae, and the latter behaviours can be used to identify a response to hypoxia in embryos which
have yet to develop the hyperventilatory response (Jonz and Nurse, 2005). Hypoxia (23 Torr)
was produced by bubbling the solution in 500 ml Nalgene bottles with nitrogen gas and verified
using an oxygen meter. Control solution (150 Torr) was kept in a separate plastic bottle. The
tubing to connect the perfusate to the perfusion chamber was gas impermeable (Tygon, Saint-
Gobain Performance Plastics Corporation, Pittsburgh PA). Ventilation frequency measurements
were conducted about 3 minutes after the fish were placed in the observation chamber to allow
the fish to recover from transfer. Fish were observed for 30 seconds and the frequency of
22
movement was recorded manually using a hand-held counter and stopwatch (Fisher Scientific).
Values were expressed as the number of gill beats or body movements per minute. Ventilation
frequency measurements of fish treated with 6-OHDA and those acclimated to hypoxia and
hyperoxia, as described above, were completed in the same manner.
2.6 Morphometric analysis
To determine whether the abundance or morphology of serotonergic skin NECs changed
during early development, or following acclimation to various levels of water PO2 (section 2.3),
parameters such as cell number, number per tissue area (density), and size (two-dimensional
area) were recorded by hand through the use of the software ImageJ (National Institutes of
Health). Once tissue was labelled with the 5-HT antibody (see section 2.2), images of the head,
trunk and tail were collected for each individual sample, using a microscope equipped with
epifluorescence (Axiophot (Zeiss or Nikon). Cells were identified on each sample area (i.e.
head, trunk, and tail), quantified using the “cell counter plugin” to keep track of the number of
cells, and presented as an average of all samples for each developmental stage. Density of cells
was calculated as the number of cells per unit area (e.g. 200 μm2) for each sample area, and then
presented as the average density per mm2.
The two-dimensional area of cells through development was also calculated using ImageJ.
Once images were collected, the clearest picture (usually the tail region) for each individual was
converted into a black-and-white (“threshold”) image, and area was automatically calculated by
the software ( . At random, five cells were selected per individual and presented as the
average of all samples for each developmental stage.
23
Distance between varicosities and points of contact for the 6-OHDA treatments (section 2.4)
was also conducted using the above software. Varicosities are swellings along a nerve axon that
represent contact with another cell or fibre. The distance between varicosities, therefore is a
measure of varicosity density. Points of contact reflect putative synapses with NECs. Oil
immersion images of skin NECs and zn-12-positive nerve fibres were collected using a confocal
microscope (see section 2.2). Points of contact between a NEC and nerve fibres were
determined from images as any apparent overlap of a nerve fibre varicosity on the surface of the
cell. The distance between varicosities was determined from images of zn-12 positive nerve
fibres. A grid was placed over each image, so that the measurements could be taken from the
same grid regions on each image. A line was hand drawn between two varicosities in any given
region, and the length of the line was determined by the software.
2.7 Statistics
Data are presented as mean ± S.E.M. for skin NEC number, density, area, and ventilation
frequencies. Multiple comparisons within one treatment group were performed using ANOVA
followed by Bonferroni test. Comparisons between different treatment groups were performed
using two-way ANOVA followed by Bonferroni test. Varicosity distance and points of contact
for 6-OHDA treated zebrafish were compared to control values using a paired t-test. Statistical
analyses were performed using commercial software (GraphPad Prism v.5.03; GraphPad
Software Inc., San Diego, CA).
24
3. RESULTS
3.1 General features of skin neuroepithelial cells (NECs)
In all zebrafish embryos and larvae observed, the presence of serotonergic skin NECs was
determined by indirect immunohistochemistry using a primary antibody directed against
serotonin (5-HT) and a secondary antibody conjugated with fluorescein isothiocyanate (FITC),
causing the cells to appear green in colour. As shown in the representative images of Figure 1,
NECs were distributed on the entire surface of the skin in larval zebrafish, including the head,
eyes, trunk, yolk sac and tail. Cell bodies appeared round in shape and were approximately 10
µm in diameter. Cells were not clustered together, but instead found in isolation. Pieces of skin
taken from adult (3 month old) zebrafish showed a drastic reduction in the number of observable
cells using 5-HT immunolabelling (Figure 1D, E). Skin NECs of adults were observed primarily
on the tail, with very few to none observed on the trunk.
Skin NECs were associated with nerve fibres in the skin of zebrafish. As Figure 2
demonstrates with cells located in the tail, skin NECs appeared to be closely associated with
nerve fibres positively labelled with the zn-12 antibody. There was a high density of zn-12
positive nerve fibres in the skin (Fig. 2B, E) and these appeared to surround and make contact
with the skin NECs (Fig. 2C, F). These same associations were present with all skin NECs
regardless of location.
25
Figure 1. Distribution of serotonergic (5-HT)-positive skin NECs of larval and adult zebrafish
as labelled with an antibody against 5-HT. A-C: Skin NECs of a 5 day post-fertilization (d.p.f.)
zebrafish. Skin NECs were distributed all over the skin surface, including the head (A), trunk
(B), and tail (C). D-E: Pieces of skin from a 3 month old adult zebrafish as labelled with anti-5-
HT. Skin NECs were not found in trunk skin (D), but were sparsely seen in tail skin (E). A skin
NEC is indicated with an arrow in each panel. F: Schematic diagram of larval zebrafish showing
location of images in A-C. Scale bar = 50 μm in A-C; 100 μm in D-E.
26
A B
C D E
F
A
B C
5-HT
27
Figure 2. Innervation of skin NECs with zn-12-positive nerve fibres. A-C: Distribution of 5-
HT positive skin NECs (A) and zn-12 positive nerve fibres (B) of the tail in a 5 d.p.f. larva. C:
5-HT and zn-12 labelling in A and B shown together revealed that skin NECs were intimately
associated with nerve fibres. D-F: Higher magnification of areas outlined in A-C of skin NECs
and associated nerve fibres. D: One 5-HT positive skin NEC. E: zn-12 labelled nerve fibres. F:
Panels D and E shown together show nerve associations with skin NEC. Scale bar = 100 μm in
A-C; 5 μm in D-F.
28
5-HT
5-HT
zn-12
zn-12
5-HT
zn-12
5-HT
zn-12
A B C
D E F
29
3.2 Development of skin NECs and effects of water PO2
To determine when skin NECs and their nervous innervation first appeared in developing
zebrafish, embryos from 24-48 hours post-fertilization (h.p.f.) were examined after co-labelling
with 5-HT and zn-12 antibodies. 5-HT immunolabelling indicated that NECs were not present at
24 h.p.f., although zn-12-immunoreactive nerve fibres had already begun to invade regions
where NECs would later be found (Figure 3A, B). At 26-28 h.p.f., NECs were numerous in the
skin and were predominately associated with nerve fibres (Figure 3C-F). NECs continued to
increase in number until 48 h.p.f. Therefore, these observations indicated that nerve fibres
appear in the skin first and NECs arise between 24 and 26 h.p.f.
Morphological parameters of skin NECs, such as number, density per tissue area and size
(i.e. two-dimensional projection area) were obtained in larvae at different developmental stages
and under different values of water PO2. Figure 4 displays the change in number and density of
NECs under control conditions (150 Torr), mild and severe hypoxia (80 and 30 Torr) and
hyperoxia (300 Torr). Interestingly, under control conditions the number and density of NECs
decreased over time, with the lowest number and density occurring at 7 d.p.f. (59±6 cells and
262±25 cells mm-2
, respectively, Fig. 4A). A similar decrease occurred when fish were reared
under chronic mild hypoxia (80 Torr), with the lowest number of cells occurring at 9 d.p.f. (70±9
cells, Figure 4B). A more severe level of hypoxia (30 Torr) elicited no significant decrease of
skin NECs occurring at any developmental stage. No significant changes in density were evident
under chronic mild (Fig. 4B) and severe hypoxic (Fig. 4C) treatment. When reared under
chronic hyperoxia (300 Torr), the number of NECs was markedly lower at 3 d.p.f. and were
further reduced at 5 d.p.f. (p<0.01; paired t-test); whereas no change, with respect to density, was
evident over development (Fig. 4D). As summarized in Figure 4E, these data collectively show
30
a significant effect of water PO2 on NEC populations during early development, such that the
developmental loss of skin NECs was reduced or delayed by chronic hypoxia (7 and 9 d.p.f.,
p<0.01; two-way ANOVA-Bonferroni test) and accelerated by chronic hyperoxia (3 and 5 d.p.f.,
p<0.01; two-way ANOVA-Bonferroni test), when compared to fish reared under control
conditions. Compared to control conditions, as shown in Figure 4F the density of skin NECs
was significantly higher during all stages of development during chronic, severe hypoxia
exposure, including 3 d.p.f. (p<0.001; two-way ANOVA-Bonferroni test). This notable
difference is likely related to an overall decrease in body size due to hypoxia, since absolute
number remains stable and the fish is not increasing in length as normally observed in normoxia
(see Appendix I). As with cell number, the density of NECs is significantly decreased when fish
were chronically exposed to hyperoxic conditions (3 and 5 d.p.f., p<0.001; two-way ANOVA-
Bonferroni test).
Two-dimensional area of skin NECs, a measure of cell size, was generally consistent
throughout development in controls and larvae exposed to mild or severe hypoxia (Figure 5A-C).
Cell area ranged between 31 and 33 µm2 under these conditions. However, the area of skin
NECs decreased significantly during chronic exposure of fish to chronic hyperoxia, producing
areas as low as 6.2±0.5 µm2 (p<0.01; ANOVA-Bonferroni test; Fig. 5D). As summarized in
Figure 5E, skin NECs were smaller at 3 d.p.f. during the 30 Torr exposure (p<0.01; ANOVA-
Bonferroni test) and from 3-7 d.p.f. during 300 Torr exposure (p<0.01; ANOVA-Bonferroni
test), compared to controls.
31
Figure 3. Skin-NECs and their innervation in embryonic zebrafish as labelled with 5-HT and
zn-12 antibodies. A: 24 hours post-fertilization (h.p.f.) zebrafish tail with no observable skin-
NECs. B: 24 h.p.f. zebrafish tail showing zn-12 nerve fibre labelling of image in A. C: 26 h.p.f.
zebrafish tail with numerous skin-NECs (one indicated with arrow). D: zn-12 labelling of image
in E showing association of nerve fibres with most skin-NECs (arrow). E: 28 h.p.f. zebrafish tail
with numerous skin-NECs (arrow). F: image in G showing innervation of skin-NECs (arrow).
Scale bar =100 µm and applies to panels A-F.
32
.
26 h.p.f.
28 h.p.f.
C D
E F
5-HT
5-HT
5-HT zn-12
5-HT zn-12
26 h.p.f.
28 h.p.f.
A B
5-HT 5-HT zn-12
24 h.p.f. 24 h.p.f.
26 h.p.f.
28 h.p.f.
33
Figure 4. The number and density (mm-2
) of NECs found on the skin of developing zebrafish
(3-9 d.p.f.) is dependent on developmental stage and water PO2. A: Under control rearing
conditions (PO2 = 150 Torr) the number (closed circles) and density (open circles) of skin NECs
decreased with time. B: Similarly to A, number and density decreased over time in zebrafish
subject to chronic exposure to mild hypoxic (PO2 = 80 Torr) rearing conditions. C: Severe
hypoxic rearing conditions (PO2 = 30 Torr) resulted in no change of NEC number and density
during the stages tested. D: Chronic exposure to hyperoxia (PO2 = 300 Torr) led to a more rapid
decline in both number and density of skin NECs. Asterisks in panels A-D indicate significant
difference from 3 d.p.f. (A-C: p<0.01; ANOVA-Bonferroni test, D: p<0.01; paired-t-test). Mean
(± S.E.M) are indicated. E: Line graph summarizes data in panels A-D of skin NECs under
hypoxic (30 Torr) and hyperoxic (300 Torr) rearing conditions, respectively, compared to control
conditions. Asterisks indicate significant difference from control (p<0.01; two-way ANOVA-
Bonferroni test). F: Line graph summarizes density data in panels A-D of skin NEC density
under hypoxic (30 Torr) and hyperoxic (300 Torr) rearing conditions, respectively, compared to
control conditions. Asterisks indicate significant difference from control (p<0.001; two-way
ANOVA-Bonferroni test). Sample size ≥8 in all cases, except for hyperoxia at 7 d.p.f., where
n=1.
34
0
60
120
180
0
100
200
300
400
500Control
Number
Density
Num
ber
5-H
T c
ells
Den
sity 5
-HT
cells (m
m-2)
0
60
120
180
0
100
200
300
40080 Torr
Density
Number
Num
ber
5-H
T c
ells
Den
sity 5
-HT
cells (m
m-2)
3 5 7 90
60
120
180
0
200
400
600
80030 Torr
Density
Number
Developmental Stage (d.p.f.)
Nu
mb
er 5
-HT
cel
ls
Den
sity 5
-HT
cells (m
m-2)
3 5 7 90
60
120
180
0
200
400
600
800300 Torr
Density
Number
Developmental Stage (d.p.f.)
Nu
mb
er 5
-HT
cel
ls
Den
sity 5
-HT
cells (m
m-2)
3 5 7 90
60
120
180
Control
80 Torr
30 Torr
300 Torr
Developmental Stage (d.p.f.)
Nu
mb
er 5
-HT
cel
ls
3 5 7 90
200
400
600
800
Control
80 Torr
30 Torr
300 Torr
Developmental Stage (d.p.f.)
Den
sity
5-H
T c
ells
(m
m-2
)
*
* *
**
A
**
B
*
**
* *
**
**
* *
C D
E F
35
Figure 5. The size of skin NECs is affected by chronic hyperoxia but not chronic hypoxia in 3-9
d.p.f. zebrafish. Projection area of skin NECs chronically exposed to (A) normoxia (PO2 = 150
Torr), (B) mild hypoxia (PO2 = 80 Torr), (C) severe hypoxia (PO2 = 30 Torr) and (D) hyperoxia
(PO2 = 300 Torr). Mean (± S.E.M) are indicated. Area of cells decreased over time in hyperoxia
(D; asterisk indicates significant difference from 3 d.p.f. (p<0.01; ANOVA-Bonferroni test). E:
Line graph summarizing data from panels A-D showing significant difference from control with
asterisks (p<0.01; two-way ANOVA-Bonferroni test). Sample size ≥ 25 in all cases, except for
hyperoxia at 7 d.p.f., where n = 5.
36
3 5 7 90
10
20
30
40
30 Torr
Developmental Stage (d.p.f.)
Are
a 5
-HT
cel
ls (
m2)
0
10
20
30
40
Control
Are
a 5
-HT
cel
ls (
m2)
0
10
20
30
40
80 Torr
Are
a 5
-HT
cel
ls (
m2)
3 5 7 90
10
20
30
40
300 Torr
Developmental Stage (d.p.f.)
Are
a 5
-HT
cel
ls (
m2)
3 5 7 90
10
20
30
40
30 Torr
80 Torr
Control
300 Torr
Developmental Stage (d.p.f.)
Are
a 5
-HT
cel
ls (
m2)
*
*
* *
*
A B
C D
E
37
3.3 Development of the hypoxic response
Zebrafish embryos at 1 d.p.f. display spontaneous whole-body movements whose frequency
was decreased from 14±2 min-1
to 1±0.4 min-1
at 2 d.p.f. (Figure 6A). During these early stages,
the frequency of whole-body movement was increased by application of hypoxia at 2 d.p.f. to
8±2 min-1
but not at 1 d.p.f. (Fig. 6A). This behaviour has been discussed in previous work
(Jonz and Nurse, 2005), and is thought to improve cutaneous gas exchange in developing
embryos (Rombough, 1988). The response to hypoxia at 3 d.p.f. and later was measured as an
increase of the frequency of buccal or opercular movements (hyperventilation) which were not
observed at earlier stages. Hyperventilation was maintained for the rest of the developmental
stages observed, as shown in Figure 6B, and reached its peak of 168±13 min-1
at 5 d.p.f.
While hyperventilation is a typical response to hypoxic stimuli in the zebrafish and many
other species of fish (Perry et al., 2009), this response did not occur when zebrafish were raised
under conditions in which water PO2 was changed, as shown in Figure 7. Chronic exposure of
zebrafish to either hyperoxia (300 Torr; Fig. 7B) or hypoxia (30 Torr; Fig. 7C) resulted in
development of a hypoventilatory response (or decrease in ventilation frequency) when exposed
to acute hypoxia. The most prominent changes in ventilation frequency observed were from
36±8 min-1
to 1.3±1.1 min-1
in larvae pre-exposed to hyperoxia (Fig. 7B) and from 40±10 min-1
to 10.5±4 min-1
in larvae pre-exposed to hypoxia (Fig. 7C). As well, a shift in peak basal
ventilation was evident under these rearing conditions, as shown in Figure 7A-C and summarized
in Figure 7D. Under control conditions, a peak basal ventilation frequency of 61±7 min-1
was
first observed at 4 d.p.f., and maintained at this approximate level until 6 d.p.f. (p>0.05;
ANOVA-Bonferroni test) before decreasing in magnitude. Under hyperoxic rearing conditions,
peak basal ventilation frequency shifted to 4 d.p.f. and was reduced to 36±8 min-1
(Fig. 7B, D).
38
Conversely, under hypoxic rearing conditions, the peak basal ventilation frequency shifted to a
later stage at 6 d.p.f. and was reduced to 40±10 min-1
(Fig. 7C, D). These shifts in peak basal
ventilation frequency in acclimated larvae, and the onset of a hypoventilatory response to acute
hypoxia (rather than a hyperventilatory response), may suggest underlying morphological
changes of O2 chemoreceptors. Interestingly, shifts in peak basal ventilation corresponded to
changes in the number and density of skin NECs reared under these same conditions, where
hypoxic-acclimated zebrafish displayed more skin NECs with a delayed development of peak
basal ventilation frequency, while hyperoxic-acclimated zebrafish displayed a rapid loss of skin
NECs and a rapid development of peak basal ventilation (recall section 3.2 and Fig. 4).
3.4 Innervation of skin NECs by catecholaminergic nerve fibres
It was hypothesized that if the skin NECs share a similar role in O2 chemoreception as gill
NECs, their innervation pattern would be similar. Gill NECs have been shown to be innervated
by at least three different nerve fibre types, including those that are catecholaminergic and of
spinal (sympathetic) origin (Jonz and Zaccone, 2009). Therefore the effects of the
catecholaminergic nerve toxin, 6-hydroxydopamine (6-OHDA), were determined to see if this
treatment would cause degeneration of nerves of the skin that make contact with serotonergic
NECs (Figure 8). The nerve density of zn-12-positive fibres was greatly reduced in zebrafish
treated with 250 μM 6-OHDA for 5 days (Fig. 8A, B). Upon closer inspection of nerve fibre
varicosities (presumed regions of contact with other cells or axons), it was found that the
distance between nerve varicosities (an inverse measure of the density of synaptic contacts along
a single nerve fibre) was significantly increased with 6-OHDA treatment, and this effect was
dependent on 6-OHDA concentration and exposure time (Fig. 8C, D, G). Moreover, compared
39
Figure 6. Development of behavioural responses to hypoxia in zebrafish from 1-9 d.p.f. A:
Measurement of whole body movements indicated that fish first respond to hypoxia at 2 d.p.f.
B: Increased ventilation in response to hypoxia is evident at 4 d.p.f. and continues for the rest of
development. Acute hypoxia exposure PO2 = 23 Torr. Sample sizes indicated in the figure.
Asterisks indicate significant difference from control (p<0.05; two-way ANOVA-Bonferroni
test).
40
3 4 5 6 7 8 90
50
100
150
200Normoxia
Hypoxia
Developmental Stage (d.p.f.)
Ven
tila
tio
n F
req
uen
cy
(m
in-1
)
(25)
(25)
(26)
(26)
(18)
(18)
(20)
(20)
(21)
(21)
(10)
(10)
(12)
(12)
*
**
**
B
1 20
5
10
15
20
25Normoxia
Hypoxia
Developmental Stage (d.p.f.)
Wh
ole
bo
dy
mo
vem
ent
(min
-1)
(31)
(31)
(33)
(33)*
A
41
Figure 7. Chronic exposure of zebrafish to hypoxia and hyperoxia caused a hypoventilatory
response to hypoxia and a shift in peak basal ventilation frequency. A: Zebrafish raised in
normoxic conditions (PO2 = 150 Torr) hyperventilated when exposed to acute hypoxia (data
from Fig. 6B). B: When raised in hyperoxic conditions (PO2 = 300 Torr), zebrafish
hypoventilated when exposed to acute hypoxia. Note also the shift in peak basal ventilation from
5 d.p.f (in A) to 4 d.p.f. C: Pre-exposure to hypoxic conditions (PO2 = 30 Torr) resulted in a
hypoventilatory response to acute hypoxia. Note also the shift in peak basal ventilation from 5
d.p.f. (in A) to 6 d.p.f. Acute hypoxia PO2 = 23 Torr. Sample size ≥ 18 for all treatments and
stages. Asterisks indicate significant difference from normoxia (p<0.05; two-way ANOVA –
Bonferroni test). D: Summary of data from A-C illustrates that peak basal ventilation is shifted
to a later developmental stage (6 d.p.f.) when fish were raised in hypoxic conditions, whereas
chronic exposure to hyperoxia shifted the peak to an earlier developmental stage (4 d.p.f.),
compared to normoxic rearing conditions.
42
0
50
100
150
200
Normoxia
Hypoxia
Ven
tila
tio
n F
req
uen
cy
(m
in-1
)
0
10
20
30
40
50
Normoxia
Hypoxia
pre-exposed to 300 Torr
Ven
tila
tion F
requen
cy (
min
-1)
0
20
40
60
Normoxia
Hypoxia
pre-exposed to 30 Torr
Developmental Stage (d.p.f.)
Ven
tila
tion F
requen
cy (
min
-1)
3 4 5 6 70
20
40
60
80
Normoxia
30 Torr
300 Torr
Developmental Stage (d.p.f.)
Ven
tila
tion F
requen
cy (
min
-1)
*
**
*
*
*
A
B
C
D
43
to controls, the number of synaptic contacts between NECs and zn-12-positive nerve fibre
varicosities was diminished in zebrafish treated with 6-OHDA (Fig. 8E, F, H). These data
indicate that the skin and, more specifically skin NECs, received catecholaminergic innervation.
Since not all zn-12-positive nerve fibres and varicosities had degenerated following 6-OHDA
treatment, this suggests that the serotonergic skin cells receive additional multiple innervation by
non-catecholaminergic nerve terminals that are not sensitive to 6-OHDA, as do NECs of the gill.
Behavioural experiments conducted on zebrafish after prolonged treatment with 6-OHDA
showed a change in both basal and hypoxia-induced ventilation frequency (Figure 9). When
chronically exposed to 6-OHDA, basal ventilation frequency was significantly reduced
compared to control from 4-6 d.p.f. (Fig. 9A; p<0.05; two-way ANOVA-Bonferroni test),
suggesting that the ventilatory drive that may normally be provided by peripheral
chemoreceptors was being suppressed by 6-OHDA-induced nerve terminal degradation. When
further stimulated with acute hypoxia, 6-OHDA treated fish were unable to hyperventilate, as
shown in Figure 9B, and the response to hypoxia was significantly reduced when compared to
that of untreated controls (p<0.001; two-way ANOVA-Bonferroni test). These data suggest that
the effects of 6-OHDA may be on innervation of peripheral chemoreceptors in the skin.
An alternative explanation to these data is that 6-OHDA had other negative effects directly on
the animals’ capacity to hyperventilate. To rule out the putative effects of 6-OHDA treatment on
the motor ventilatory reflex, zebrafish were acutely exposed to NaCN, a potent ventilatory
stimulant, to artificially induce hyperventilation. As Figure 9C demonstrates, both control and 6-
OHDA treated zebrafish tested at 5 d.p.f. responded equally well to NaCN in the form of
hyperventilation (p<0.01; two-way ANOVA-Bonferroni test). These results confirm that 6-
OHDA treatment does not directly impair the ability of larvae to hyperventilate.
44
Figure 8. 6-hydroxydopamine (6-OHDA) exposure leads to neuronal loss in the skin of
developing zebrafish and loss of contact between skin NECs and zn-12 positive nerve fibres. A-
B: Compared to control, 6-OHDA treated fish had lost zn-12-positive nerve fibres, as shown in
the tail of this 7 d.p.f. larva. Larvae were treated for 5 days with 6-OHDA at a concentration of
250 µM. C-D: Compared to control, 6-OHDA treated fish had a longer distance between
varicosities, suggesting reduced varicosity density. C: 5 d.p.f. control. D: 7 d.p.f. treated for 5
days with 6-OHDA at a concentration of 250 µM. E-F: Images of skin NEC with associated
nerve fibres. E: 5 d.p.f. control. F: 7 d.p.f. larva treated for 5 days with 6-OHDA at a
concentration of 250 µM showed reduced points of contact between nerve fibres and NEC. G:
Distance between varicosities was significantly increased by 3 day treatment of 500 μM 6-
OHDA and 5 day treatment of 250 μM 6-OHDA. Sample sizes indicated in the figure. Asterisk
indicates significant difference from control (p<0.05; paired t-test). H: Points of contact of
nerve varicosities or terminals on skin NECs was significantly reduced only by application of
250 μM 6-OHDA for 3 days. Sample sizes indicated in the figure. Asterisk indicates significant
difference from control (p<0.05; paired t-test). Scale bar = 100 µm in A and B; 5 µm in C and
D; 5 µm in E and F.
45
A B
C D
E F
Control
Control
Control
6-OHDA
6-OHDA
6-OHDA
zn-12 zn-12
zn-12zn-12
5-HT zn-12 5-HT zn-12
5 70
1
2
3
4
5Control
6-OHDA (250M)
6-OHDA (500 M)
Developmental Stage (d.p.f.)D
ista
nce
bet
wee
n v
aric
osi
ties
(
m)
5 70
2
4
6
8Control
6-OHDA (250 M)
6-OHDA (500 M)
Developmental Stage (d.p.f.)
Po
ints
of
co
nta
ct
G
H
(30)(30) (30)
(30)
(15)*
*
(4)
(4)(6)
(3) (3)
*
46
Figure 9. 6-OHDA treatment reduced basal ventilation frequency and eliminated response to
acute hypoxia. A: Basal ventilation frequency was reduced in fish treated with 250 μM 6-
OHDA compared to control (normoxia) from 4-6 d.p.f.. B: Zebrafish did not hyperventilate
when exposed to acute hypoxia after 6-OHDA treatment. Hyperventilation in control fish (see
Fig. 6B) when exposed to hypoxia is shown as a trend line for comparison. C: 6-OHDA treated
zebrafish (5 d.p.f.) hyperventilated when exposed to sodium cyanide (NaCN). Sample sizes
indicated in the figure. Asterisks in A and C indicate significant difference from control (A:
p<0.05; C: p<0.01; two-way ANOVA-Bonferroni test).
47
3 4 5 6 70
20
40
60
80Normoxia
Normoxia (6-OHDA)
Developmental Stage (d.p.f.)
Ven
tila
tion F
requen
cy (
min
-1)
3 4 5 6 70
50
100
150
200
Normoxia (6-OHDA)
Hypoxia (6-OHDA)
Hypoxia
Developmental Stage (d.p.f.)
Ven
tila
tion F
requen
cy (
min
-1)
Control 6-OHDA0
50
100
150Normoxia
NaCN
Chronic Treatment
Ven
tila
tion F
requen
cy (
min
-1)
A
B
C
(25) (66)
(26)
(72)
(18)
(50)
(20)
(30)
(21)(20)
(9)
(8) (8)
(9)
* *
*
*
*
48
4. DISCUSSION
This study focused on characterizing neuroepithelial cells (NECs) found on the skin of
developing zebrafish. Although skin NECs were briefly noted in one study as putative O2
chemoreceptors (Jonz and Nurse, 2006), their detailed morphology and potential role in
development have not been established. The purpose of the current thesis, therefore, was to
further describe the morphology, distribution and development of skin NECs in embryos and
larvae, and examine their putative role in O2 sensing. Figure 10 summarizes the results of this
study.
4.1 Skin NECs as oxygen chemoreceptors
There exists evidence of chemosensory cells in the skin of both developing and adult fish,
which include taste buds (TBs) and solitary chemoreceptor cells (SCCs; Kotrschal et al., 1997;
Hansen et al., 2002), so it is reasonable to propose that other chemosensory cells (i.e. O2 sensing
cells) exist in the skin. Compared to characteristics of TBs and SCCs, the skin NECs appear to
be a distinct population of cells, not like the other surface chemoreceptors found on the zebrafish
skin. All three cell types can be found on the entire body surface, with the distribution of TBs
being somewhat limited to the head in most species. There is an even distribution of skin NECs
and SCCs over the body surface; however, the cells in the present study have a higher density on
the body and head compared to the tail, which differs from the density distribution of SCCs.
While the skin NECs decrease in number and density as the zebrafish develops, SCCs and cells
of the TBs increase with development (Kotrschal et al., 1997; Hansen et al., 2002). Therefore,
these notable differences define the skin NECs as a newly described surface chemoreceptor.
49
Figure 10. Chronology of the development of ventilatory events and structures in zebrafish.
Solid arrows correspond to results obtained from the present study and dashed arrows indicate
work of other authors. 1Kimmel et al., 1995;
2Rombough, 2002;
3Higashijima et al., 2000;
4Jonz
and Nurse, 2005. d.p.f., days post-fertilization; NECs, neuroepithelial cells; HV,
hyperventilatory.
50
d.p.f 1 2 3 4 5 6 7 8 14
• Skin-NECs appear
• Skin-NECs receive
innervation
• Pectoral fin
response to hypoxia
• Max # and
density of
skin-NECs
• Max resting vent.
(Chronic Hyperoxia)
• HV response to
hypoxia
• Max HV
response to
hypoxia
• Max
resting vent.
(Chronic
Hypoxia)
• Min # and density
of skin-NECs
• Hatch1
• Gills ventilated2
• Gill arches
innervated3
•HV response to
hypoxia4
• NECs of gill
filaments appear
and innervation
begins4
• All gill
NECs
innervated4
• Lamellae
begin to form4
• Buccal
movements regular4
• Branchial respiration begins2
51
Characterizing skin NECs as oxygen chemoreceptors requires a comparison between these
cells and existing identifiable oxygen chemoreceptors, such as NECs of the fish gill,
neuroepithelial bodies (NEBs) of the mammalian lung, carotid body (CB) type I cells, and
adrenomedullary chromaffin cells (AMCs). NEBs, NECs, AMCs and type I cells function as
oxygen chemoreceptors, releasing neurotransmitters to produce appropriate sensory responses,
such as hyperventilation, to hypoxic challenges (Cutz and Jackson, 1999; Lopez-Barneo et al.,
2008; Nurse et al., 2009; Perry et al., 2009). All of these O2 chemoreceptors share
commonalities with the currently studied skin NECs. The presence of the neurotransmitter
serotonin is an important indicator of some oxygen chemoreceptors (e.g. Cutz and Jackson,
1999; Perry et al., 2009) and is present in the cells of the present study (Figure 1). Hypoxia
evokes serotonin secretion in mammalian NEBs, which produces local vascular responses (Cutz
and Jackson, 1999; Fu et al., 2002). In addition, serotonin release has also been shown to be
important for vascular changes in the fish gill (Sundin et al., 1995). The role of serotonin as a
neurotransmitter involved in oxygen sensing in the gill, however, has yet to be established.
Each type of peripheral O2 chemoreceptor is also in contact with sensory nerve fibres in order
to transmit nerve signals associated with hypoxic stimulation to the central nervous system, and
skin NECs also appeared to make associations with nerve fibres (Figure 2). As previously
discussed (see sections 1.2.2, 1.2.3), one type of sensory nerve fibre CB type I cells and gill
NECs establish contact with is of the catecholaminergic type. Utilizing the neurotoxin 6-
hydroxydopamine (6-OHDA), it was demonstrated that the skin of zebrafish contained
catecholaminergic innervation, and skin NECs also established connections with these
catecholaminergic nerve fibres (Figure 8). However, not all nerve terminals in contact with skin
NECs were destroyed by 6-OHDA (Figure 8F), suggesting that these cells make contact with
52
other types of nerve fibres. Other types of nerve fibres which establish contact with other
oxygen chemosensory cells, such as parasympathetic efferent nerves, afferent serotonergic
nerves and efferent nitrergic nerves (Campanucci and Nurse, 2007; Jonz and Zaccone, 2009)
may also establish contact with skin NECs. However, characterization of such additional
innervation will require further study.
Skin NECs were present and appeared innervated well before 2 days post-fertilization (d.p.f.)
in developing zebrafish (Figure 3). This phenomenon coincided with the approximate time when
zebrafish were no longer capable of anoxia-induced metabolic supression, an indication of
anoxia tolerance (>24 hours post-fertilization, h.p.f., Padilla and Roth, 2001), and when a
hypoxic response was first observed in the form of increased pectoral fin and whole body
movements (2 d.p.f., Jonz and Nurse, 2005; Figure 6A). Innervation of the skin NECs several
hours before the embryo could actively respond to hypoxia may represent the first step in the
development of a functional chemosensory pathway. The early embryo appears to be in a
transition state from which it conforms to hypoxia (or anoxia) at 24 h.p.f. by metabolic
suppression to one in which it will respond to hypoxia at 2 d.p.f. The small size of the embryo at
these transitional stages may preclude any need for a response to hypoxia since the metabolic
requirements of oxygen uptake are met through cutaneous respiration (Rombough, 2002).
Nerve fibres positively labelled with zn-12 antibody appeared in the skin before skin NECs
were observed (Figure 3), suggesting that innervation may induce the development of skin
NECs. The neural induction model has been proposed as the mechanism for TB development
(Northcutt, 2004), although another study done in zebrafish found that innervation of TBs
occurred after TB primordia, suggesting an induction model which is independent of innervation
(Hansen et al., 2002).
53
The number and density of skin NECs gradually decreased over days 3-9 of development
(Figure 4A), reaching their lowest values at 7 d.p.f.; the same day that gill NECs are identified as
becoming functional because all of the cells receive sensory innervation (Jonz and Nurse, 2005).
The decrease of skin NECs over development suggests that a functional role for these cells may
be confined to early developmental stages (i.e. before 7 d.p.f.), especially since skin NEC
populations are markedly decreased in 3 month old adults (Figure 1D, E), although whether the
skin NECs are active or not in adults is unknown. As the gills become functional for both
respiration and O2 sensing later in larval development, zebrafish no longer rely solely on the skin
for respiration (Rombough, 2002). Thus, the reduction in number and density of skin NECs
during this time may be related to this shift in respiration from the skin to gills. As demonstrated
in the mammalian model, the AMCs are highly sensitive to hypoxia during the first few weeks of
life and become developmentally insensitive as the CB type I cells receive innervation and
become mature (Nurse et al., 2009); thus, it is not unreasonable to propose the same mechanism
could hold true for the zebrafish. The skin NECs may serve as the early oxygen sensors until gill
NECs are mature and take over as the adult oxygen chemoreceptors.
Exposing zebrafish to chronic hypoxia and hyperoxia during the first 9 days of development
affected the number and density of skin NECs (Figure 4B-F). Chronic hypoxia has been said to
be a teratogen and negatively affect fish development (Shang and Wu, 2004; Wu, 2009).
However, the number and density of skin NECs did not significantly change when zebrafish
were raised in severe hypoxia (PO2= 30 Torr), compared to the developmental loss of these cells
in controls. This relative increase in both number and density with severe hypoxic treatment
indicates that the number of skin NECs may be upregulated during chronic hypoxia, a response
typical of O2 chemoreceptors (Nurse and Vollmer, 1997; Wang and Bisgard, 2002; Jonz et al.,
54
2004). It is unknown whether this upregulation is due to a reduction in cell death, or an increase
in cell division. Furthermore, it is interesting to note that chronic hypoxia appears to slow the
development of gill NECs in zebrafish beyond 7 d.p.f. during early larval stages (Shakarchi and
Jonz, unpublished observations), so the persistence of more skin NECs (with putative O2
chemosensitivity) during this time may be beneficial. Chronic hypoxia appeared to slow body
growth, and produced developmental tachycardia (see Appendix I), but it did not appear to
impede the stability of the skin NEC population, thereby strengthening the conclusion that
hypoxia was having a specific effect on these cells. The two-dimensional projection area of skin
NECs during hypoxic exposure (Figure 5B, C) was unaffected, indicating that hypoxia does not
induce hypertrophy in skin NECs. This is in contrast to acclimation studies of other oxygen
chemoreceptors, as hypoxia usually led to an increase in area (Jonz et al., 2004; Nurse and
Vollmer, 1997; Wang and Bisgard, 2002).
In mammals, chronic exposure to elevated oxygen leads to a reduction in size of the CB,
fewer CB type I cells, and an attenuation of the CB’s response to hypoxia (Bisgard et al., 2005).
Hyperoxia has also been demonstrated to induce oxidative cell death (Horowitz, 1999). In adult
zebrafish, the density of gill NECs decreased after chronic exposure to hyperoxic conditions
(Vulesevic et al., 2006). Likewise, upon chronic hyperoxic exposure, the skin NEC population
was severely reduced (Figure 4D) and many fish did not survive past 7 d.p.f. (N.B an n=1 at 7
d.p.f. and no values for 9 d.p.f. in Figure 4D). The reduction of skin NECs may be due to
hyperoxia- induced oxidative cell death. As well, the two dimensional area of skin NECs was
reduced after hyperoxic exposure (Figure 5D). Therefore, hyperoxia had a similar effect on skin
NECs as it does on other oxygen chemoreceptors.
55
4.2 The hypoxic ventilatory response
Consistent with previous studies, zebrafish began to respond to hypoxia at 2 d.p.f. by
increasing whole body and pectoral fin movements (Figure 6A; Jonz and Nurse, 2005). This
early response to hypoxia has been observed in embryonic and larval fish and may facilitate gas
exchange. Gill ventilation activity has been demonstrated to be coordinated with pectoral fin
movements in larvae, and movement of the pectoral fin may improve gas exchange. This
behaviour can, thus, be used to identify a response to hypoxia before the hyperventilatory
response develops (Jonz and Nurse, 2005). Early zebrafish motor behaviour includes
spontaneous side-to-side contractions beginning at 17 h.p.f., which peak in frequency by 19 h.p.f.
and then decline progressively over the course of 6–7 hours (Drapeau et al., 2002). At hatching
(2 d.p.f.) the swimming contraction frequency increases, but the larvae swim only infrequently
and in bursts (Drapeau et al., 2002). Thus, the increased level of movement seen in the
normoxic exposed fish at 1 d.p.f. (Figure 6A) may be explained by these frequent, spontaneous
movements, which then decline by 2 d.p.f. Increased ventilation frequency in response to
hypoxia begins at 3 d.p.f. and continues over the rest of development (Figure 6B). In contrast to
a previous study, peak hyperventilation in response to hypoxia occurred at 5 d.p.f instead of 7
d.p.f. (Jonz and Nurse, 2005). This may be due in part to the decreased PO2 used in this study
compared to the other (23 Torr versus 25 Torr), or the chronological staging used in this study
compared to the staging used in the other study following Kimmel et al. (1993).
While hyperventilation is the typical response to hypoxic stimuli in the zebrafish (Figure 6)
and many other species of fish (Perry et al., 2009), this response can be reversed under different
rearing conditions, as seen in Figure 7B, C. Chronic exposure of zebrafish to either hypoxic or
hyperoxic conditions resulted in a hypoventilatory response (or decrease in ventilation
56
frequency) when exposed to acute hypoxia. Exposing zebrafish for the first 7 days of
development to hyperoxia has been shown to alter adult reflexes to ventilatory stimuli including
hypoxia, whereas chronic exposure to hypoxia does not alter these reflexes (Vulesevic and Perry,
2006). This study reasoned that there is a developmental window during maturation (i.e. the
first 7 days of development) when the potential for respiratory plasticity is high. Blunting of
ventilation to acute hypoxia in adults also occurs when developing mammals are exposed to
chronic hypoxia or hyperoxia (Lahiri et al., 2000; Bisgard et al., 2005). Early insults to the fish
(i.e. hypoxic or hyperoxic conditions) can disrupt the normal maturation of respiratory control,
and these insults may be evident during this critical period of development. The fish may then
be unable to hyperventilate when faced with acute hypoxia. This may be due to a reduced O2
sensitivity of skin NECs, similarly to the reduced O2 sensitivity of the CB in this same situation,
or a change in innervation during these exposures, which has been reported in the CB (Carroll,
2003).
A shift in peak basal ventilation was evident under hypoxic and hyperoxic rearing conditions,
as shown in Figure 7C. Under control conditions, the peak basal ventilation frequency was first
observed at 4 d.p.f., and was maintained until 7 d.p.f. Under hyperoxic rearing conditions, the
peak basal ventilation frequency shifted to 4 d.p.f. Conversely, under hypoxic rearing
conditions, the peak basal ventilation frequency shifted to a later stage, 6 d.p.f. Interestingly,
these shifts corresponded to changes seen in the population of skin NECs under these same
rearing conditions (Figure 4). The number and density of cells were rapidly reduced, even at the
earliest stages observed, when zebrafish were reared under hyperoxic conditions. During chronic
hypoxia exposures, the skin NEC population did not significantly change over development and
remained present in abundance at later developmental stages. This lends credence to the role of
57
skin NECs in sensing oxygen and playing a role in driving basal ventilation. In mammals,
peripheral chemoreceptors have been shown to play an important role in maintaining the resting
level of ventilation (Gonzalez et al., 1994). It has been found in fish, however, that peripheral
chemoafferent input does not appear to be essential, but may provide some input, for resting
rhythmic ventilation (Smatresk, 1990).
Upon treatment with 6-OHDA, basal ventilation frequency was decreased and the response to
hypoxia was eliminated in developing zebrafish (Figure 9A, B). As shown in Figure 8F, there
was a loss of nerve varicosities in contact with skin NECs, which may have contributed to the
ventilation frequency changes. The CB has been shown to be important in maintaining the basic
resting operation of breathing pattern generation in the brain stem, and denervation of the carotid
sinus nerve led to changes in resting respiration patterns (Gonzalez et al., 1994). Thus, the loss
of skin NEC catecholaminergic innervation may have suppressed resting ventilation frequency,
and this may have contributed to the loss of the hypoxic ventilatory response. 6-OHDA,
however, may not completely eliminate the ability of developing zebrafish to sense decreased
O2, as treatment with sodium cyanide (NaCN) induced hyperventilation (Figure 9C). NaCN
halts cellular respiration by inhibiting cytochrome c oxidase in mitochondria, essentially
mimicking tissue and cellular anoxia (Morocco, 2005). Thus, acute stimulation by NaCN may
have induced a lower intracellular PO2 in these experiments, possibly even anoxia, compared to
acute application of hypoxic solution at a PO2 of 23 Torr. These potent effects of NaCN may
explain the hyperventilatory behaviour observed.
58
5. CONCLUSIONS AND FUTURE DIRECTIONS
The main objective of this thesis was to answer the question as to whether skin NECs are
oxygen sensitive chemoreceptors during early development in zebrafish. The present results
show that skin NECs are an attractive site for oxygen chemoreception as they share similar
properties to other identified oxygen chemoreceptors, such as catecholaminergic innervation,
retention of serotonin and display changes in cell number, and density when exposed to hypoxia
and hyperoxia. Skin NECs were present and innervated before 2 d.p.f. when zebrafish first
displayed behavioural responses to acute hypoxia, and the cells steadily declined over time,
reaching a minimum at the time when gill NECs become functional. Pre-exposure of fish to
hypoxic and hyperoxic conditions changed the way larvae reacted to acute hypoxia exposure.
Skin NECs may also play a role in driving resting ventilation frequency. This thesis provides an
in depth description of skin NECs as being extrabranchial oxygen chemoreceptors in early
developing zebrafish.
Physiological examination during hypoxia exposures would elucidate the role of skin NECs
as early oxygen chemoreceptors. As well, further investigation is needed to determine what
other types of innervation are associated with these cells, and what is triggering developmental
disappearance: the appearance of gill NECs, or another mechanism. It would also be of interest
to investigate as to whether these cells are present in any other vertebrates. By creating a profile
of serotonergic skin NECs as being oxygen chemoreceptors, this thesis contributes to the
understanding of the evolutionary and developmental aspects of oxygen sensing in vertebrates in
general. Given what is known about changes in O2 sensitive sites in mammals with
development, from AMCs in neonates to CB type I cells in adults, this research suggests that a
59
transition from one O2 sensing site to another with development may be a phenomenon that arose
early in vertebrate phylogenesis.
60
6. REFERENCES
Bailly Y, Dunel-Erb S, Laurent P. 1992. The neuroepithelial cells of the fish gill filament:
indolamine-immunocytochemistry and innervation. Anat. Rec. 233: 143-161.
Bamford OS. 1974. Oxygen reception in the rainbow trout (Salmo gairdneri). Comp. Biochem.
Physiol. 48A: 69-76.
Barnard P, Andronikou S, Smatresk N, Mokashi A, Lahiri S. 1987. Time-dependent effect of
hypoxia on carotid body chemosensory function. J. Appl. Physiol. 63: 685-691.
Barrionuevo WR, Burggren WW. 1999. O2 consumption and heart rate in developing zebrafish
(Danio rerio): influence of temperature and ambient O2. Am. J. Physiol. Regulatory
Integrative Comp. Physiol. 276: 505-513.
Bisgard GE. 2000. Carotid body mechanisms in acclimation to hypoxia. Respir. Physiol. 121:
237-246.
Bisgard GE, Olson EB, Bavis RW, Wenninger J, Nordheim EV, Mitchell GS. 2005. Carotid
chemoafferent plasticity in adult rats following developmental hyperoxia. Respir.
Physiol. Neurobiol. 145: 3-11.
Blum D, Torch S, Lamberg N, Nissou M-F, Benabid A-L, Sadoul R, Verna J-M. 2001.
Molecular pathways involved in the neurotoxicity of 6-OHDA, dopamine and MPTP:
contribution to the apoptotic theory in Parkinson’s disease. Prog. Neurobiol. 65:
135-172.
Bradford CS, Sun L, Collodi P, Barnes DW. 1994. Cell cultures from zebrafish embryos and
adult tissues. Mol. Mar. Biol. Biotechnol. 3: 78-66.
Burleson ML, Milson WK. 2003. Comparative aspects of O2 chemoreception: anatomy,
physiology, and environmental adaptations. In: S Lahiri, GL Semenza, NR Prabhakar
(eds.) Oxygen Sensing: Responses and Adaptation to Hypoxia. pp. 685-707. New York:
Marcel Dekker.
Campanucci VA, Nurse CA. 2007. Autonomic innervation of the carotid body: role in efferent
inhibition. Respir. Physiol. Neurobiol. 157: 83-92.
Carroll J. 2003. Developmental plasticity in respiratory control. J. Appl. Physiol. 94: 375-389.
Coccimiglio ML. 2008. Characterization of serotonergic epithelial cells of the skin in
developing zebrafish. B.Sc. Honours Thesis, pp. 47.
Cutz E, Jackson A. 1999. Neuroepithelial bodies as airway oxygen sensors. Respir. Physiol.
115: 201-214.
61
Diaz RJ, Breitburg DL. 2009. The hypoxic environment. In JG Richards, AP Farrell, CJ
Brauner (eds.) Hypoxia. pp. 1-23. London, UK: Elsevier Inc.
Døving KB, Kasumyan AO. 2008. Chemoreception. In: RN Finn, BG Kapoor (eds.) Fish
Larval Physiology. pp. 331-393. Enfield, NH, USA: Science Publishers.
Drapeau P, Saint-Amant L, Buss RR, Chong M, McDearmid JR, Brustein E. 2002.
Development of the locomotor network in zebrafish. Prog. Neurobiol. 68(2): 85-111.
Dunel-Erb S, Bailly Y, Laurent P. 1982. Neuroepithelial cells in fish gill primary lamellae. J.
Appl. Phyiol. 56: 1342-1353.
Finley JCW, Polak J, Katz DM. 1992. Transmitter diversity in carotid body afferent neurons:
dopaminergic and peptidergic phenotypes. Neurosci. 51(4): 973-987.
Fu XW, Nurse CA, Wong V, Cutz E. 2002. Hypoxia-induced secretion of serotonin from intact
pulmonary neuroepithelial bodies in neonatal rabbit. J. Physiol. 539(2): 503-510.
González C, Almaraz L, Obeso A, Rigual R. 1994. Carotid body chemoreceptors: from natural
stimuli to sensory discharges. Physiol. Rev. 74: 829-898.
Hansen A, Reutter K. 2004. Chemosensory Systems in Fish: Structural, Functional and
Ecological Aspects. In: G von der Emde, J Mogdans, BG Kapoor (eds.) The Senses of
Fish: Adaptations for the Reception of Natural Stimuli. pp. 55-89. New Delhi, India:
Dordrecht, Netherlands.
Hansen A, Reutter K, Zeiske E. 2002. Taste bud development in the zebrafish, Danio rerio.
Dev. Dyn. 223: 483-496.
Higashijima SI, Hotta Y, Okamoto H. 2000. Visualization of cranial motor neurons in live
transgenic zebrafish expressing green fluorescent protein under the control of the Islet-l
promoter/enhancer. J. Neurosci. 20: 206-218.
Horowitz S. 1999. Pathways to cell death in hyperoxia. Chest 116: 64S-67S.
Jonz MG, Nurse CA. 2003. Neuroepithelial cells and associated innervation of the zebrafish
gill: a confocal immunofluorescence study. J. Comp. Neurol. 461: 1-17.
Jonz MG, Fearon IM, Nurse CA. 2004. Neuroepithelial oxygen chemoreceptors of the zebrafish
gill. J. Physiol. 560.3: 737-752.
Jonz MG, Nurse CA. 2005. Development of oxygen sensing in the gills of zebrafish. J. Exp.
Biol. 208: 1537-1549.
Jonz MG, Nurse CA. 2006. Ontogenesis of oxygen chemoreception in aquatic vertebrates.
Resp. Physiol. Neurobiol. 154: 139-152.
62
Jonz MG, Zaccone G. 2009. Nervous control of the gills. Acta histochem. 111: 207-216.
Katz DM, Markey KA, Goldstein M, Black IB. 1983. Expression of catecholaminergic
characteristics by primary sensory neurons in the normal adult rat in vivo. Proc. Natl.
Acad. Sci. USA 80: 3526-3530.
Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling TF. 1995. Stages of embryonic
development of the zebrafish. Dev. Dyn. 203: 253-310.
Kotrschal K. 1991. Solitary chemosensory cells – taste, common chemical sense or what? Rev
Fish Biol. Fisheries 1: 3-22.
Kotrschal K, Krautgartner W-D, Hansen A. 1997. Ontogeny of the solitary chemosensory cells
in the zebrafish, Danio rerio. Chem. Senses 22: 111-118.
Lahiri S, Roy A, Baby SM, Hoshi T, Semenza GL, Prabhakar NR. 2006. Oxygen sensing in the
body. Prog. Biophys. Mol. Biol. 91: 249-286.
Lahiri S, Rozanov C, Cherniack NS. 2000. Altered structure and function of the carotid body at
high altitude and associated chemoreflexes. High Altitude Med. Biol. 1: 63-74.
Laurent P, Rouzeau J-D. 1972. Afferent neural activity from pseudobranch of the teleosts.
Effects of PO2, pH, osmotic pressure and Na+ ions. Respir. Physiol. 14: 307-331.
Le Guellec D, Morvan-Dubois G, Sire J-Y. 2004. Skin development in bony fish with particular
emphasis on collagen deposition in the dermis of the zebrafish (Danio rerio). Int. J. Dev.
Biol. 48: 217-231.
López-Barneo J, López- López JR, Ureña J, González C. 1988. Chemotransduction in the
carotid body: K+ current modulated by PO2 in type I chemorecptor cells. Science 241:
580-582.
López-Barneo J, Nurse CA, Nilsson GE, Buck LT, Gassmann M, Bogdanova AY. 2010. First
aid kit for hypoxic survival: sensors and strategies. Physiol. Biochem. Zool. 83(5):
753-763.
López-Barneo J, Ortega-Sáenz P, Pardel R, Pascual A, Piruat JI. 2008. Carotid body oxygen
sensing. Eur. Respir. J. 32: 1386-1398.
McDonald DM, Mitchell RA. 1975. The innervation of glomus cells, ganglion cells and blood
vessels in the rat carotid body: a quantitative ultrastructural analysis. J. Neurocytol. 4:
177-230.
Mendelsohn BA, Kassebaum BL, Gitlin JD. 2008. The zebrafish embryo as a dynamic model of
anoxia tolerance. Dev. Dyn. 237: 1780-1788.
63
Metcalfe WK, Myers PZ, Trevarrow B, Bass MB, Kimmel CB. 1990. Primary neurons that
express the L2/HNK-1 carbohydrate during early development in the zebrafish. Dev.
110: 491-504.
Milsom WK, Reid SG, Rantin FT, Sundin L. 2002. Extrabranchial chemoreceptors involved in
respiratory reflexes in the neotropical fish Colossoma macropomum (the tambaqui). J.
Exp. Biol. 205: 1765-1774.
Mills E, Jöbsis FF. 1972. Mitochondrial respiratory chain of carotid body and chemoreceptor
response to changes in oxygen tension. J. Neurophysiol. 35: 405-428.
Mills L, Nurse C. 1993. Chronic hypoxia in vitro increases volume of dissociated carotid body
chemoreceptors. Neuroreport 4: 619-622.
Morocco AP. 2005. Cyanides. Crit. Care Clin. 21: 691-705.
Northcutt RG. 2004. Taste buds: development and evolution. Brain Behav. Evol. 64: 198-206.
Nurse CA, Buttigeig J, Brown S, Holloway AC. 2009. Regulation of oxygen sensitivity in
adrenal chromaffin cells. Ann. NY Acad. Sci. 1177: 132-139.
Nurse CA, Vollmer C. 1997. Role of basic FGF and oxygen in control of proliferation, survival,
and neuronal differentiation in carotid body chromaffin cells. Dev. Biol. 184: 197-206.
Ortega-Saenz P, Garcia-Fernandez M, Pardal R, Alvarez E, López-Barneo J. 2003. Studies on
glomus cell sensitivity to hypoxia in carotid body slices. Adv. Exp. Med. Biol.
536: 65-73.
Padilla PA, Roth MB. 2001. Oxygen deprivation causes suspended animation in the zebrafish
embryo. Proc. Natl. Acad. Sci. USA. 98: 7331-7335.
Parng C, Roy NM, Ton C, Lin Y, McGrath P. 2007. Neurotoxicity assessment using zebreafish.
J. Pharmocol. Toxicol. Methods. 55: 103-112.
Parry G, Holliday FGT. 1960. An experimental analysis of the function of the pseudobranch in
teleosts. J. Exp. Biol. 37: 344-354.
Perry SF, Gilmour KM. 2002. Sensing and transfer of respiratory gases at the fish gill. J. Exp.
Zool. 293: 249-263.
Perry SF, Jonz MG, Gilmour KM. 2009. Oxygen sensing and the hypoxic ventilatory response.
In JG Richards, AP Farrell, CJ Brauner (eds.) Hypoxia. pp. 193-253. London, UK:
Elsevier Inc.
64
Podrasky JE, Lopez JP, Fan TWM, Higashi R, Somero GN. 2007. Extreme anoxia tolerance in
the embryos of the annual killifish Austrofundulus limnaeus: insights from a
metabolomics analysis. J. Exp. Biol. 210: 2253-2266.
Qin Z, Lewis JE, Perry SF. 2010. Zebrafish (Danio rerio) gill neuroepithelial cells are sensitive
chemoreceptors for environmental CO2. J. Physiol. 588.5: 861-872.
Rico AJ, Prieto-Lloret J, Gonzalez C, Rigual R. 2005. Hypoxia and acidosis increase the
secretion of catecholamines in the neonatal rat adrenal medulla: an in vitro study. Am. J.
Physiol. Cell Physiol. 289: C1417-C1425.
Rombough P. 1988. Respiratory gas exchange, aerobic metabolism, and effects of hypoxia
during early life. In WS Hoar, DJ Randall (eds.) Fish Physiology. pp 59-161. New
York, London: Academic Press.
Rombough P. 2002. Gills are needed for ionoregulation before they are needed for O2 uptake in
developing zebrafish, Danio rerio. J. Exp. Biol. 205: 1787-1794.
Saunders RL, Sutterlin AM. 1971. Cardiac and respiratory responses to hypoxia in the sea
raven, Hemitripterus americanus, and an investigation of possible control mechanisms.
J. Fish Res. Bd. Canada 28: 491-503.
Schwerte T. 2009. Cardio-respiratory control during early development in the model animal
zebrafish. Acta histochem. 111: 230-243.
Shang EHH, Wu RSS. 2004. Aquatic hypoxia is a teratogen and affects fish embryonic
development. Environ. Sci. Technol. 38: 4763-4767.
Slotkin TA, Seidler FJ. 1988. Adrenomedullary catecholamine release in the fetus and
newborn: secretory mechanisms and their role in stress and survival. J. Dev. Physiol.
10: 1-16.
Smatresk NJ. 1990. Chemoreceptor modulation of endogenous respiratory rhythms in
vertebrates. Am. J. Physiol. 259: R887-R897.
Smith PG, Mills E. 1976. Autoradiographic identification of the terminations of petrosal
ganglion neurons in the cat carotid body. Brain Res. 113: 174-178.
Stea A, Jackson A, Nurse CA. 1992. Hypoxia and N6,O
2’-dibutyryladenosine 3’,5’-cyclic
monophosphate, but not nerve growth factor induce Na+ channels and hypertrophy in
chromaffin-like arterial chemoreceptors. Proc. Natl. Acad. Sci. USA 89: 9469-9473.
Sundin L, Nilsson GE, Block M, Löfman O. 1995. Control of gill filament blood flow by
serotonin in the rainbow trout, Oncorhynchus mykiss. Am. J. Physiol. 268: R1224-
R1229.
65
Thompson RJ, Jackson A, Nurse CA. 1997. Developmental loss of hypoxic chemosensitivity in
rat adrenomedullary chromaffin cells. J. Physiol. 498.2: 503-510.
Trevarrow B, Marks DL, Kimmel CB. 1990. Organization of hindbrain segments in the
zebrafish embryo. Neuron. 4: 669-679.
Turesson J, Schwerte T, Sundin L. 2006. Late onset of NMDA receptor-mediated ventilatory
control during early development in zebrafish (Danio rerio). Comp. Biochem. Physiol.
143A: 332-339.
Turesson J, Sundin L. 2003. N-methyl-D-aspartate receptors mediate chemoflexes in the
shorthorn sculpin Myoxocephalus scorpius. J. Exp. Biol. 206: 1251-1259.
Vulesevic B, McNeill B, Perry SF. 2006. Chemoreceptor plasticity and respiratory acclimation
in the zebrafish Danio rerio. J. Exp. Biol. 209: 1261-1273.
Vulesevic B, Perry SF. 2006. Developmental plasticity of ventilatory control in zebrafish,
Danio rerio. Respir. Physiol. Neurobiol. 154: 396-405.
Wang Z-Y, Bisgard GE. 2002. Chronic hypoxia-induced morphological and neurochemical
changes in the carotid body. Micro. Res. Tech. 59: 168-177.
Westerfield M. 2000. The zebrafish book. A guide for the laboratory use of zebrafish (Danio
rerio), 4th
ed. Eugene, OR: University of Oregon Press.
Whitear M. 1952. The innervation of the skin of teleost fishes. Q. J. Microsc. Sci. 93: 298-305.
Whitear M. 1971. The free nerve endings in fish epidermis. J. Zool., Lond. 163: 231-236.
Whitear M. 1992. Solitary chemosensory cells. In: TJ Hara (ed.) Fish Chemoreception, 2nd
ed.
pp. 103-125. London: Chapman and Hall.
Wu RSS. 2009. Effects of hypoxia on fish reproduction and development. In JG Richards, AP
Farrell, CJ Brauner (eds.) Hypoxia. pp. 79-141. London, UK: Elsevier Inc.
Zaccone G, Ainis L, Mauceri A, Lo Cascio P, Lo Giudice F, Fasulo S. NANC nerves in the
respiratory air sac and branchial vasculature of the Indian catfish, Heteropneustes fossilis.
Acta histochem. 105: 151-163.
66
APPENDIX I
Measurements of heart rate frequency and body length were completed under a dissecting
microscope (Leica MZ6). Body length (from snout to tail) was measured to the nearest 0.1 mm.
Because early larvae are translucent, heart rate frequency was measured visually through the
body wall. Heart rate was observed in individuals over one 30 second time interval (value
expressed per minute) and were recorded using a hand-held counter and stopwatch (Fisher
Scientific). Fish were transferred to a glass bottomed culture dish using a glass pipette and were
lightly anaesthetized with 0.05 mg/ml MS222 (as described in section 2.5). Measurements were
made 1-2 minutes after transfer.
67
Figure 1. Length and heart rate of developing zebrafish during chronic hypoxia exposure. (A)
length of zebrafish is negatively affected up until 7 d.p.f. when chronically exposed to severe
hypoxia (30 Torr). Sample sizes indicated in the figure. Asterisks indicate significant difference
from control means (p<0.05; two-way ANOVA-Bonferroni test). (B) Heart rate of zebrafish
when chronically exposed to severe hypoxia (30 Torr) shows a reduction at 5 d.p.f., and then an
increase at 9 d.p.f. Sample sizes indicated in the figure. Asterisks indicate significant difference
from control means (p<0.001; two-way ANOVA-Bonferroni test).
68
3 5 7 90
1
2
3
4
5Control
Hypoxia (30 Torr)
Developmental Stage (d.p.f.)
Len
gth
(m
m)
3 5 7 90
50
100
150Control
Hypoxia (30 Torr)
Developmental Stage (d.p.f.)
Hea
rt R
ate
(min
-1)
(11)
(14)
(15)
(16)
(11)
(14)
(15)
(16)
(14)
(14)
(20)(20)
(20)
(20)(20)
(20)
**
*
**
A
B
top related